ࡱ>   ŏbjbj  '_bb&|!'X5(5(5($Y(Y(Y(P(,), Y(l79(:::@b7ItK<,i5(Lw@^@LLy'y'::h $QQQLy':5(:QLQQ;{8'6: |2= Y(sO.s}*0<l}jOT##>5(a,LLQLLLLLQLLLlLLLLLLLLLLLLLb k%:  Outcomes from combining workand tertiary study Cain Polidano Rezida Zakirova Melbourne Institute of Applied Economic and Social Research The views and opinions expressed in this document are those of the author/project team and do not necessarily reflect the views of the Australian Government, state and territory governments or ϲʿ¼. Any interpretation of data is the responsibility of the author/project team. Publishers note To find other material of interest, search VOCED (the UNESCO/ϲʿ¼ international database < HYPERLINK "http://www.voced.edu.au/" http://www.voced.edu.au>) using the following keywords: transition from education and training to employment; training employment relationship; work experience; employability; employment; youth; higher education; vocational education and training; tertiary education.  Commonwealth of Australia, 2011 This work has been produced by the ϲʿ¼ (ϲʿ¼) under the National Vocational Education and Training Research and Evaluation (NVETRE) Program, which is coordinated and managed by ϲʿ¼ on behalf of the Australian Government and state and territory governments. Funding is provided through the Department of Education, Employment and Workplace Relations. Apart from any use permitted under the Copyright Act 1968, no part of this publication may be reproduced by any process without written permission. Requests should be made to ϲʿ¼. The NVETRE program is based upon priorities approved by ministers with responsibility for vocational education and training (VET). This research aims to improve policy and practice in the VET sector. For further information about the program go to the ϲʿ¼ website . The author/project team was funded to undertake this research via a grant under the NVETRE program. These grants are awarded to organisations through a competitive process, in which ϲʿ¼ does not participate. The views and opinions expressed in this document are those of the author/project team and do not necessarily reflect the views of the Australian Government, state and territory governments or ϲʿ¼. ISBN 978 1 921809 47 7 web edition 978 1 921809 48 4 print edition TD/TNC 102.29 Published by ϲʿ¼ ABN 87 007 967 311 Level 11, 33 King William Street, Adelaide, SA 5000 PO Box 8288 Station Arcade, Adelaide SA 5000, Australia ph +61 8 8230 8400 fax +61 8 8212 3436 email ncver@ncver.edu.au About the research Outcomes from combining work and tertiary study Cain Polidano and Rezida Zakirova, Melbourne Institute of Applied Economic and Social Research Working in some capacity is almost considered de rigueur for tertiary students. The reasons for working and the impact this has on both an individuals ability to complete their studies and on their post-study labour market outcomes are only recently receiving attention. Using the 1995 and 1998 cohorts of the Longitudinal Surveys of Australian Youth (LSAY), this study investigates the motivations for and the education and employment outcomes from working while studying for both vocational education and training (VET) and higher education students. The authors find that income is an important motivating factor: those in receipt of income support are less likely to work while studying, although this is dependent on whether the student is still living at home. Key messages For those studying full-time, working impacts on completionthe more hours worked, the greater the effect. For example, working 1624 hours a week reduces the completion rate by eight percentage points, while more than 24 hours reduces it by 14percentage points. Finding work in a job considered a career job while studying has a significant and positive impact on course completion for both VET and higher education students. For all tertiary students, being employed in the final year of study improves the chances of finding full-time employment, even three years after completing the course. Interestingly, for both full- and part-time students, the longer they have been employed in a job, the greater the likelihood of course completion, while past work experience also increases the likelihood of completion for full-time students (2.5 percentage points per year of employment). Perhaps this reflects that these students have better time management skills. Thus it is clear that combining study and work does have significant effects on completion and future employment prospects. Too much work negatively impacts on study completion, but on the other hand work experience does benefit future job prospects. The ideal combination would be modest hours of work in a job relevant to a future careerbut this will be difficult to achieve for many students. Tom Karmel Managing Director, ϲʿ¼ Contents  TOC \o "1-2" Tables  PAGEREF _Toc151540212 \h 6 Executive summary  PAGEREF _Toc151540213 \h 7 Introduction  PAGEREF _Toc151540218 \h 11 Literature to date  PAGEREF _Toc151540219 \h 12 Employment benefits of combining work with study  PAGEREF _Toc151540220 \h 12 Impacts on education outcomes  PAGEREF _Toc151540221 \h 13 Data and sample  PAGEREF _Toc151540222 \h 15 The sample  PAGEREF _Toc151540223 \h 15 Student jobs  PAGEREF _Toc151540224 \h 17 Course outcomes  PAGEREF _Toc151540225 \h 20 Post-study employment outcomes  PAGEREF _Toc151540226 \h 23 Modelling approach  PAGEREF _Toc151540227 \h 25 Modelling the choice of study and work combinations  PAGEREF _Toc151540228 \h 25 Modelling education outcomes  PAGEREF _Toc151540229 \h 27 Modelling employment benefits from combining work andstudy  PAGEREF _Toc151540230 \h 28 Results  PAGEREF _Toc151540231 \h 29 Choice of work and study combinations  PAGEREF _Toc151540232 \h 29 Student outcomes  PAGEREF _Toc151540233 \h 32 Employment outcomes  PAGEREF _Toc151540234 \h 36 Conclusions  PAGEREF _Toc151540235 \h 38 References  PAGEREF _Toc151540236 \h 40 Appendix  PAGEREF _Toc151540237 \h 41  Tables  TOC \f f \t "tabletitle,1" \c "Figure" 1 Frequency of course enrolments in the sample  PAGEREF _Toc153533899 \h 16 2 Work and study combinations in the year in which the first tertiary course ended  PAGEREF _Toc153533900 \h 17 3 Whether or not job in the year in which the first tertiary course ended was a career job  PAGEREF _Toc153533901 \h 18 4 Occupation for those whose first tertiary course was VET  PAGEREF _Toc153533903 \h 19 5 Occupation for those whose first tertiary course was higher education  PAGEREF _Toc153533904 \h 20 6 Outcomes of first tertiary course  PAGEREF _Toc153533905 \h 21 7 Completion rates of first tertiary course by hours of work in the year before ending study  PAGEREF _Toc153533906 \h 22 8 Found a career job before end of first tertiary course  PAGEREF _Toc153533907 \h 22 9 Employment outcomes in the first three years after ending the final spell of study  PAGEREF _Toc153533908 \h 24 10 Marginal effects for choice of work and study combination for first tertiary course  PAGEREF _Toc153533909 \h 30 11 Marginal effects of student outcomes, all first-time tertiary courses  PAGEREF _Toc153533910 \h 32 12 Key interactions between VET and hours of work in first tertiary courses  PAGEREF _Toc153533911 \h 34 13 Predicted probabilities of full-time employment in the first three years after ending study  PAGEREF _Toc153533912 \h 36 A1 Employment rates and hours of work during higher education students first tertiary course  PAGEREF _Toc153533913 \h 41 A2 Level of VET qualification studied for those aged 1525 whose first tertiary course is a VET course  PAGEREF _Toc153533914 \h 41 A3 Coefficients for the ordered probit model of post-study employment outcomes  PAGEREF _Toc153533915 \h 42  Executive summary Background For the majority of tertiary students aged 25 and younger in Australia, working while studying is the norm; however, the motivating factors and educational and post-study labour market outcomes of working during this period of their lives are largely unknown. For young people, the chance to experience work is likely to help them to develop both general and job-specific skills that will help them successfully transit into the labour market after study. On the downside, there is also the risk that time spent in work may take away from time studying, thereby reducing chances of completion and damaging future labour market prospects. For policy-makers, understanding motivations and measuring outcomes is important for designing policies for youth that on one hand provide them with support to complete their studies but, on the other, do not diminish the benefits of work. This is the first study in Australia to use multivariate analysis to examine motivations and education and employment outcomes from working while studying for both vocational education and training (VET) students (excluding apprentices and trainees) and higher education students aged 25 years and under. It is important to try and eliminate the effects of confounding factorsfactors that are related to both working and outcomesthat distort the relationship between outcomes and work while studying. In contrast to descriptive statistics, the use of multivariate analysis allows us to determine whether any observed relationship between hours of work and course completion, for example, is due to hours of work or a third variable, such as socioeconomic background. We use the 1995 and 1998 cohorts of the Longitudinal Surveys of Australian Youth (LSAY) to undertake this work. LSAY contains detailed individual information on youth, including information on socioeconomic status and past education outcomes (including university entry scores). Another key feature of LSAY is that it tracks the same individuals from the time they are 15 (in 1995 and 1998for the two cohorts respectively) until they are 25. This longitudinal aspect is important in the context of this study because we examine employment outcomes from combining work and study upto three years after completion. Taking a longer-term view helps give a clearer picture of potential employment benefits, as it is likely that any initial benefits will diminish over time. Motivations for combining work and study Consistent with the findings of the Australian House of Representatives Standing Committee on Education and Training (2009), we find that, for those aged 25 and under in their first tertiary course, the socioeconomic status of parents (measured by employment status, job prestige and highest level of education achieved) has very little bearing on the choice of work and study combination. This suggests that students are not motivated by financial need. Instead, we find that receipt of Youth Allowance (means-tested income support for students) and culture are important factors in the choice of work and study combinations. Overseas-born youth are estimated to be ten percentage points less likely to combine work and study (either by enrolling full-time and working part-time or enrolling part-time and working) than Australian-born students. Similarly, those who receive Youth Allowance are less likely to combine work and study, but the degree of lower participation in work among Youth Allowance recipients depends on whether they are living with their parents or not. For those who live with their parents, receiving Youth Allowance reduces their likelihood of choosing to combine work and study by 20 percentage points, whereas those who do not live at home and receive Youth Allowance are 13 percentage points less likely to combine work and study. The smaller impact of Youth Allowance on the choice to combine work and study for those living away from home is likely to be because these students face higher costs of living than those living at home, which necessitates some work. Those who receive Youth Allowance are also less likely to choose work and study combinations that require a considerable commitment of time in work, such as working more than 16 hours (roughly two days) a week, and studying full-time or working full-time and studying part-time. However, we cannot conclude whether the relationships between Youth Allowance and choice of work and study combinations are causal. It is possible that students who choose more onerous work commitments prefer working longer hours because, for example, they have greater financial commitments and hence forego receiving Youth Allowance. Compared with higher education students, we find that VET students are eight percentage points less likely to combine work and study. However, those who do are seven percentage points more likely to combine work and part-time study. The preference for part-time study for VET students may be because more of them already have ongoing full-time employment, but it may also be because many have to pay up-front fees. Educational outcomes After controlling for differences between VET and higher education students, including academic ability, we find that full- and part-time VET students aged 25 and under in their first tertiary course are around ten percentage points more likely to complete than their higher education counterparts. The higher completion rates among this group of VET students may be due to a number of reasons, including the shorter duration of the courses, differences in the academic demands of the courses and differences in the flexibility and modes of course delivery. In general, the modularised nature of VET means that courses can be better tailored to individual training needs and are delivered in a greater range of modes, especially off-campus delivery modes. After controlling for differences between those who choose different work and study combinations, including academic ability, course load, courses types and field of study, we find that for those aged 25 and under in their first tertiary course, working while studying can reduce the chances of completing, but it depends on the hours worked. For full-time students, we find that compared with those who do not work while studying, those who work up to eight hours (roughly a day) a week on average while studying are just as likely to complete, while those who work more than eight hours are less likely to complete; that is, those working 8.1 to 16 hours (roughly two days) a week, 16.1 to 24 hours (roughly three days) a week and those working more than 24 hours a week are five percentage points, eight percentage points and 14 percentage points less likely to complete, respectively. For part-time students, due to the small number of observations, the only comparison is between those who work fewer than 32 hours per week (part-time workers) and those who work more than 32 hours (full-time workers). We estimate that part-time students who work full-time are around 12 percentage points less likely to complete than those who work fewer than 32 hours per week. From tests performed, we find no evidence that these results are affected by self-selection biasthe presence of unobserved factors that affect both the choice of average hours worked and course completion. We find that, generally speaking, there are no differences in the ability of full-time VET students and full-time higher education students to manage work and study. However, we find that part-time VET students who work full-time (work more than 32 hours per week on average over their course) are around 15 percentage points more likely to complete than part-time higher education students who work full-time. To complete a qualification part-time while working full-time requires considerable effort and application and the longer duration of higher education courses may make the required commitment more taxing. The relatively long commitment required to obtain a higher education qualification part-time may also mean that employers are less likely to support full-time employees who choose this education pathway. Importantly, we find that, for both VET and higher education students, the type of work performed while studying has a significant bearing on completion. Full-time students who find a job they wouldlike as a career while studying (around 12% of both VET and higher education students) are estimated to be around four percentage points more likely to complete study than those who work in a job that is not a career job, while the same effect for part-time students is around ten percentage points. A possible explanation is that most of those who find a career job while studying find work in professional jobs, especially in the areas of information technology, engineering, and architecture and building, which tend to require the attainment of a qualification for post-study employment. Therefore, the prospect of converting their jobs to ongoing employment after study may give them an added incentive to complete over those who work in non-career jobs. If this interpretation is correct, this result underlines the importance of creating more opportunities for students to gain experience working in jobs that they would like as a career. A consistent result for part- and full-time students is that the longer an individual has been in the job, the greater the chance of completion. A possible explanation is that the more established an individual is in the job, the more support they may get from their employer in the form of more flexible working hours or possibly time off work, in the case of full-time employees. Similarly for full-time students, the more years of employment experience, the greater the chance of completion. The importance of work experience may be linked to the development of soft skills, such as time management, commitment to completing a task, communication and interpersonal skills and self-esteem, which may help academic performance. Alternatively, the relationship may not be causal, but instead related to uncontrolled differences in the characteristics of students who have and have not a history of working; for example, differences in motivation. Employment outcomes Results suggest that employment in the last year of study significantly improves the chances of finding full-time employment in the first year out from study, but that, for higher education students, the magnitude of benefits depends heavily on the nature of the job while studying. Compared with those who were not working in their last year of higher education, those who were casually employed in a career job and those employed in a non-career job are estimated to be 74 percentage points and 25 percentage points more likely to be in employment in their first year out, respectively. For VET students, working in a casual career job and working in a non-career job in the last year of study is associated with a 68-percentage point and a 65-percentage point higher probability of full-time employment, respectively, in the first year out. We find that working in the last year of study also has longer-term benefits for the chances of being in full-time employment, but that these benefits diminish over time. There are a number of possible explanations for why the initial employment benefits to higher education depend more heavily on the type of work performed. First, employers of higher education graduates may not value general skills developed from working in a non-career job to the same degree as employers of VET graduates. Second, employers of higher education graduates may value general skills highly, but consider course completion as a better measure of these skills than working in non-career jobs while studying. Finally, although they report that they do not want the job as a career, VET students, by comparison with higher education students, may be more likely to derive job-specific skills that are recognised by employers from a non-career job. A note of caution when interpreting the employment benefits of combining work and study in this report: while the estimated benefits appear large, especially in the initial period, we cannot rule out the possibility that some, if not all, of the estimated benefit is due to uncontrolled factors, such as personality traits, that affect both employment while studying and employment shortly after studying. Introduction In Australia, around 70% of youth in tertiary study (aged between 20 and 24) have a part or full-time job (ABS 2008). While students report working for a range of reasons, typically they do not report working to fund basic necessities (see, for example, Abhayaratna et al. 2008; Dwyer et al. 1999; Robinson 1996). Despite the high rate of participation in work, little is known about its effects on student education and employment outcomes. Combining work with study is likely to have both positive and negative effects on the short- and longer-term outcomes of students. For example, on the positive side, skills and connections made through working while studying may increase the chances of finding post-study employment. On the negative side, working can impact on the time available for study, recreation and socialising, which may reduce the chances of completing study and hence finding employment after study. Finding the right balance between work and study is not easy and may depend on a number of individual factors, such as the time spent in work, the type of work performed, course demands, academic ability, past experience in the labour market, intended future career paths, course fees, other time commitments and the availability of other forms of financial support. The aim of this project is to contribute to the understanding of the choice of work and study combinations of vocational education and training (excluding apprentices and trainees) and higher education students and their impacts on educational and employment outcomes (which was a key recommendation of the recent House of Representatives Standing Committee on Education and Training [2009]). We use the Longitudinal Surveys of Australian Youth (LSAY) (1995 and 1998 cohorts) with multivariate regression techniques to undertake this analysis. The panel nature of LSAY allows us to examine how work and study combinations affect subsequent employment outcomes, while the richness of the data enables us to control for a range of personal differences between individuals such as socioeconomic background and academic ability which may also affect completion and employment outcomes. Findings from this study may be used to evaluate the need for, and the type of, possible government intervention. Government intervention may be necessary if combinations of work and study are not in the best interests of young people, measured as post-education employment outcomes. Young people may make poor decisions if they are unable to assess the impacts of combining work and study wisely, are coerced by employers into working too many hours, or are reliant on the income to remain in study. From a policy perspective, helping young people to find the right balance between study and work will not only benefit affected students, but will help ensure that the future labour market is well skilled, which will have flow-on benefits for future labour market participation and social inclusion. An important contribution of this paper is that it considers the outcomes from combining study and work for tertiary students; Australian studies to date have mainly focused on outcomes for secondary school students. Moreover, we consider the outcomes for VET students separately from those for higher education students. Literature to date Before examining the effect that combining work and study may have on study and employment outcomes, it is pertinent to examine briefly what work has already been undertaken in this area. Employment benefits of combining work with study When surveyed, students often give a number of reasons for working while studying but, generally speaking, the reasons are not related to financial necessity. In the main, students report that they work because it gives them financial independence from their parents, it improves their self-esteem, it improves their chances of finding work after study and they enjoy it (Abhayaratna et al. 2008; Dwyer et al. 1999; House of Representatives Standing Committee on Education and Training 2009; Robinson 1996). While there may be a range of personal benefits to students, we focus only on the employment benefits of combining work and tertiary study. In general terms, we may assume that tertiary students who work may develop both general and job-specific skills relevant to their post-study employment prospects. General skills, or employability skills, refer to broad generic work-related competences and personal attributes such as communication skills, teamwork skills, problem-solving skills etc. which are valued by employers (Sweet 2008). While general skills may benefit all students because they are relevant to all employers, job-specific skills developed by combining work with study will only benefit those who are working in a job relevant to their post-study career. The extent to which tertiary students may benefit from job-specific skills is a key issue to be addressed in this study and is measured by whether students report working in a job that they would like as a career. A limitation of this measure is that it is a binary measure of the relevance of the work to a students post-study career plans. This means that students may develop some skills relevant to their chosen career path, but still report that the job is not one that they would like as a career. Ideally, we would use a measure that asks students to evaluate, on a scale, the relevance of the job to their career plans. While it is likely that the employment benefits from combining work with study will be greater if a student can find a job that matches their post-study career interests, estimating such an effect for VET students is complicated because some may have found a career job prior to study. In some cases, VET students may choose to study to meet the training needs of their career job, which will over-inflate the estimated employment benefits of combining work with study and underestimate any employment penalty from course non-completion. Empirical evidence on the benefits of working while studying to date is limited to part-time work while at school. In the international literature, there is evidence of strong positive short- and longer-term benefits on employment and wages (Carr, Wright & Brody 1996; Meyer & Wise 1982; Ruhm 1997). In Australia, Vickers, Lamb and Hinkley (2003), used LSAY (1995 cohort) and a multivariate framework to examine the relationship between working while at school and student outcomes in the first year after completing school. Results from this study show that, of those who do not go on to further study, students who worked while in school were 46% more likely to be in employment than those who did not work. Similarly, Robinson (1996) found that students who worked part-time while in Year 11 were less likely to be unemployed in their first year out and were less likely to experience spells of unemployment. Evidence of employment benefits from working while in tertiary education is only anecdotal. In a survey of Australian VET students, Dwyer et al. (1999) found that students recognised the importance of obtaining a work record for their future employment prospects, but claimed that the on the job skills they developed were not relevant to their training. Similarly, McInnis and Hartley (2006) found evidence that some higher education students are motivated to work for job-specific employment benefits. In particular, they found evidence in competitive areas, such as commerce, that some students in their last year of study would seek more professionally oriented part-time work to try and get a foot in the door. Impacts on education outcomes In theory, time spent in work is at the expense of time spent studying and time spent in recreational activities, such as socialising with peers and on-campus activities, which can affect student performance and attachment to the course, respectively (Tinto 1993). In terms of empirical studies, there is a substantial body of work on the impacts of work on higher education student performance and course completion using longitudinal data, especially in the United States (for a review of this literature, see Riggert et al. 2006). However, the findings of these studies are inconsistent, reflecting differences in approach. In Australia, Marks et al. (2000) and Robinson (1999) both used longitudinal data and found that those who worked in a part-time job in Year 11 were just as likely to complete secondary school as those who did not work in Year 11. However, Robinson (1999) did find that those who spent more than ten hours a week in employment were slightly less likely to finish secondary school compared with those who did not work. This result is consistent with the findings of Vickers, Lamb and Hinkley (2003) who examined the effect of working while in higher education. Using a multivariate model, Vickers, Lamb and Hinkley (2003) found that, after controlling for course contact hours and field of study, working 20 hours per week or more doubles the likelihood of dropping out of higher education compared with those who do not work. However, there is no significant effect estimated for those who work fewer than 20 hours compared with those who do not work. In their multivariate analysis, Vickers, Lamb and Hinkley (2003) also included VET students in their sample alongside higher education students and found smaller impacts (compared with results with higher education students only), which led them to conclude that the impacts of work on VET completion rates may be smaller. However, they did not include an interaction term to test this hypothesis. It is important to note that these results do not control for the possibility of self-selection bias. It is important to note that these results do not control for the possibility of self-selection bias. Self-selection occurs when there are factors present that affect both the hours of work selected and the chances of completion, but are not controlled for in the regression. An example may be an individuals attachment to their course. If they are not enjoying their course, all else being equal, they have a lower chance of completing because they will not see the benefit of pursuing a career in this area. At the same time, because they do not enjoy time studying for the course, they may tend to work longer hours. In this case, without controlling for course attachment, the estimated impacts of work on course completion will be over-estimated. However, we note that this is just an example and there may be a host of other uncontrolled factors that may be related to both the choice of hours and academic performance. Therefore, we cannot be certain of the direction of any bias in the above studies but, suffice to say, any bias present will depend on the nature of the controls used. This study is the first that we are aware of which attempts to deal with the issue of self-selection in the choice of working hours in determining student outcomes. If all of the differences related to hours of work cannot be controlled for (which they cannot using available datasets), then the estimated impacts of combining work and study will be biased to some degree because differences in student outcomes will be related not only to the hours of work, but to differences in the characteristics and circumstances of those who choose various levels of work. The direction of any bias is unclear. On the one hand, if the choice of hours worked while studying is driven mainly by attachment to their course, then we may expect that the bias will result in an overestimate of the impacts of combining work and study if those with weak attachment choose to work longer hours. On the other hand, if those who are better able to cope with work and study work longer hours, then the bias will underestimate the true impacts of combining work and study. In practice, the direction and extent of any bias will depend upon the nature of controls, which is probably why the empirical results to date are inconclusive. This study is the first that we are aware of which attempts to deal with the issue of self-selection in the choice of working hours in determining student outcomes. Data and sample Given that we are interested in the effects on youth transitions of combining work and post-school study, the most suitable datasets available are the Longitudinal Surveys of Australian Youth (LSAY). These datasets track individuals for ten years, from the time they are in Year 9 until they have left study. Because the data are longitudinal, they are ideal for tracking transitions from education into employment. To take advantage of all available data and to maximise the robustness of our results, we combined the 1995 and 1998 cohorts. Only variables of interest that were common across the two datasets were merged. To control for differences in the sample that may be due to differences in the survey methodology between the two cohorts, we derive a binary identifier for use in the multivariate analysis. The sample The sample of interest in this study is all tertiary students up until they are approximately 25 years old. A tertiary student is an individual who is observed to have been enrolled in study that leads to a qualification and has stopped, regardless of whether they completed or not. If an individual reports to have stopped studying since the last interview, regardless of whether they completed or not, and currently reports to be studying a different course, two separate courses are recorded for the same student. Courses that do not lead to a nationally recognised qualification upon completion, such as short courses and modules, are not included as courses. For comparison, courses are classified as either in VET or in higher education, where VET is identified as being certificates level IIV, diploma and advanced diploma, and higher education is a bachelor degree or above. A tertiary course is identified in LSAY when an individual reports being in post-school education at the time of interview, or when they report to have participated in education in between survey periods. However, in the latter case, we do not directly observe employment details while studying. If at the point of interview an individual who was in study between survey waves reports to be working in the same job as in the previous interview, we assume that their current employment conditions are the same as those while studying. Those who undertook study after the previous interview and changed jobs are omitted from the sample (114 spells of study in one or more courses). Failure to include spells that occurred in between interview periods would be likely to bias the sample because short spells (such as spells of VET and spells that did not lead to a qualification) would be underrepresented. While we endeavour to retain as many courses as possible in the sample, we omit 571 courses for which, in the last year of study, there is missing information on: course completion; the type of qualification attained; whether study was full or part-time; employment status; or average hours worked per week. In the main, information is missing because of survey attrition and other forms of non-response. However, we note that individuals who have missing observations for variables that are not as crucial, such as field of study, are not omitted. For these variables, the effect of missing information on study and employment outcomes is dealt with in the multivariate analysis by including separate missing information categories. Table 1 Frequency of course enrolments in the sample VETHigher educationFull-timePart-timeTotalFull-timePart-timeTotalCount%Count%Count%Count%Count%Count%No. of courses19447637858132270408786499664586832257212343649126603132243082715340335575476234411024406110190201105002020000000Total1245100655100190010047751007591005534100Source: LSAY 1995 and 1998 cohorts. Overall, there are 7434 courses undertaken in our sample: 1900 are VET courses and 5534 are higher education courses (table 1). An important point of note is that 70% of VET courses and 83% of higher education courses are the first tertiary courses undertaken by students and the remainder are subsequent courses. For the purposes of this analysis, we restrict analysis on the choice and academic outcomes of work and study combinations to those students undertaking their first tertiary course. We choose the first tertiary course to put VET and higher education students on equal footing, given that most higher education students only undertake one course. When examining the employment outcomes from combining work and study, we use the study and work combinations from an individuals most recently finished spell of education in the survey to gain a truer picture of how working while in study affects post-study employment outcomes. An important issue in the context of this study is when should comparisons between students be made, given that course duration varies? When examining post-study employment outcomes, we pay most attention to the effect of work in the last period of study because it is likely to exert the most influence, but also include information on employment history. Further, we choose the last period because most VET courses typically run for between six and 18 months, which means that the last year of study is the only one available for many VET spells. For analysis on completion, categorising work while studying is more complicated because the choice to work and choice of hours may be related to the academic performance of students. For example, poor-performing students may decide to work or increase their hours to prepare themselves for a premature transit from study to the labour market. In such cases, any relationship estimated would spuriously overestimate the impacts of combining work and study. Analysis of the data (table A1 in appendix) indicates that, while the average hours of work do not vary greatly throughout the duration of a course, there is an increase in the rate of work with years in study, especially for courses that stretch beyond two years (mostly higher education courses). To deal with the changing rate of employment over time, we choose to examine the effect of the average hours of work performed per week over the duration of study on course completion. This treatment captures the combined effect of working hours and the duration of work. Student jobs In terms of work and study combinations (table 2), for the first tertiary course undertaken after leaving school it is clear that VET is more likely than higher education to be undertaken part-time (35% versus 13%). Around half of part-time VET and higher education students are employed full-time (32 hours or more) in their first course. For full-time study, 37% of VET students undertake thecourse without working, compared with 23% for higher education students. Ignoring differences in the rates of work for full-time students, for first-time courses, there are only minor differences in the relative frequencies of work and study combinations between VET and higher education. Table 2 Work and study combinations in the year in which the first tertiary course ended VETHigher educationFull-timePart-timeTotalFull-timePart-timeTotalCount%Count%Count%Count%Count%Count%Not working3493758154073194523418986220.18 hrs114121131251063116204651148.116 hrs2152329824419117929601212392716.124 hrs12213339155126371669147061624.132 hrs72847131199358984174421032.140 hrs53613937192152095140283498>40 hrs121591671591282171734Total937100376100131310040501004961004546100Source: LSAY 1995 and 1998 cohorts. As discussed above, a key challenge in examining student outcomes from combining work and study is that, in many cases, student jobs are career jobs that were obtained prior to study. Students who have already found a job desired as a career before they enter study are likely to have different incentives for completing education and have an advantage in securing post-study employment. Using LSAY, we identified those individuals who reported to have had a job as a career in the year before commencing a spell of study and who had the same job in the year in which they ended their course (including those who changed to another job desired as a career while studying). Also identified are those who had a career job in the last year of study, but who did not have a career job in the year prior to commencing study. It may be assumed that these individuals found a career job during the course of their study and the job may be attributed as a benefit of undertaking the course. Data for the last year of study in the first tertiary course and all subsequent courses are presented in table 3. Table 3 Whether or not job in the year in which the first tertiary course ended was a career job VETHigher educationFull-timePart-timeTotalFull-timePart-timeTotalCount%Count%Count%Count%Count%Count%Doesn't have a career job50084187586877527318729464302584Had a job as a career before study2244514677492123612Started career job while studying7312882816118362121523351414Total59510032010091510031421004581003600100Source: LSAY 1995 and 1998 cohorts. As expected, part-time VET students are much more likely to have found a job as a career before undertaking their first course of study. For the first tertiary course undertaken, around 14% of part-time VET students already had a job as a career before commencing study, compared with 3% of part-time higher education students. Part-time students are also much more likely than full-time students to have found a job they would like as a career while studying. Around one-third of part-time higher education students and 28% of part-time VET students report to have found a job they would like as a career while studying, compared with 12% for both VET and higher education students studying full-time. While the information in table 3 identifies whether students report that their job is one they would like as a career, tables 4 and 5 present information on the types of jobs students perform across fields of study (see box 1 for a description of how the occupation categories were derived). An important point from tables 4 and 5 is that, in general, jobs performed are more strongly related to the fields of study than the course type (VET or higher education). For students undertaking their first tertiary course, around 60% of both VET and higher education students are employed in clerical, sales or services jobs in their last year before ending study. While most students work in jobs unrelated to their field of study, those employed in professional jobs (management; science and engineering; business and information technology; health; education; and arts) are more likely to be students undertaking courses in the related fields of education than not. For a given field of study, there are differences in occupations between the two groups, which are likely to be related to differences in the types of education and training and, hence, differences in the types of jobs the two groups are likely to perform post-study. A clear demonstration of this is the difference in jobs performed by VET and higher education students who undertook courses in the information technology and engineering fields of study. Around 34% of VET students who did courses in the information technology field and 37% of VET students who did courses in the engineering field were employed in production jobs (especially intermediate production), compared with 17% and 16% of higher education students from these fields. Instead of working in production, higher education students from these fields were more likely to be employed as professionals. Box 1 Deriving occupation categories in LSAYOccupation data in LSAY is coded using three different job classifications: Australian Standard Classification of Occupations (ASCO), first edition; ASCO, second edition; and Australian and New Zealand Standard Classification of Occupations (ANZSCO), first edition. For the purpose of this report, occupation classifications were produced at the ASCO second edition 2-digit level using ABS concordance (ABS 1997, 2006). Matching the ASCO first edition and ANZSCO first edition to the ASCO second edition was done using 6-digit level concordance. However, because the data available from LSAY are only available at the 4-digit level, there is no perfect fit between each of the occupations across the three classification types. Therefore, a considerable amount of judgment and trial and error was required to match the occupations at the 2-digit level. This problem was particularly problematic for converting ASCO first edition classifications (first five waves of the 1995 cohort) to the ASCO second edition. To test the validity of the match for the ASCO first edition, we compared the frequency of the matched 2-digit categories for wave 5 of the 1995 cohort with the frequency of the 2-digit ASCO second edition categories for wave 6 of the 1995 cohort. We found some significant discrepancies, especially between the intermediate and elementary workers, which are most likely to be related to problems with the match than year-on-year compositional changes. For the purposes of comparing types of jobs performed by VET and higher education students, we judged the discrepancies to be benign.Table 4 Occupation for those whose first tertiary course was VET OccupationMan.Sci. & Eng.Bus. & ITHealthEduc.ArtsCleric.ServiceSalesLab. & prod.OtherCount%%%%%%%%%%%Field of educationNat. & phys. sci.0303373173030330Info. tech.13120051672034376Eng. & rel. tech.5127020522137943Arch. & bld.9505005018411822Agr. env. & rel.3300036221641632Health2012039392913294Education000017006717006Man. & com.507003211434132238Soc. & cult.401111361836164117Arts50004651636235110Hosp. & pers. serv.5020005433010586Mixed1300000132550008Count37103259409417026316834862Note: Man = Manager and administrator, generalist manager, specialist manager, farmer and farm manager, managing supervisors; Sci. & eng. = Science, building and engineering professionals and associate professionals; Bus. & IT = Business and information professionals and associate professionals; Health = Health and welfare professionals and associate professionals; Ed. = Education professionals; Arts = Social, arts and miscellaneous professionals and associate professionals; Cleric. = Secretaries and personal assistants, other advanced, intermediate and elementary clerical workers; Service = Intermediate and elementary service workers; Sales = Intermediate and elementary sales and related workers; Lab. & prod. = Intermediate plant and machine operators, transport workers and other intermediate production and transport workers; Other = Mechanical, fabrication, automotive, electrical, construction and other tradespeople and skilled agricultural workers and other associate professionals. Source: LSAY 1995 and 1998 cohorts. Table 5 Occupation for those whose first tertiary course was higher education OccupationMan.Sci. & eng.Bus. & ITHealthEduc.ArtsCleric.ServiceSalesLab. & prod.OtherCount%%%%%%%%%%%Field of educationNat. & phys. sci.55306781534142433Info. tech.631906315721174184Eng. & rel. tech.4191004251422164195Arch. & bld.8176001010142210363Agr. env. & rel.3900228112038966Health31132563632101474Education51106362837121268Man. & com.601802416143091658Soc. & cult61404615193492813Arts614048102030153344Hosp. & pers. serv.50005717152722241Mixed10000000000001Count188101244181361823966861088427743540Source: LSAY 1995 and 1998 cohorts. Course outcomes A key focus of this study is the relationship between work and study combinations and study outcomes. Course outcomes are observed in the survey after students cease study in a particular course (up to 12 months after completion). These outcomes are completed study, deferred or withdrew (table 6). Defining completion is considered a difficult concept for VET students because many undertake a VET course simply to complete a course module and not to attain a qualification (McMillan, Rothman & Wernert 2005). In some waves in LSAY we can identify whether on enrolment an individual was working towards the completion of a course or had enrolled simply to complete a module. From the data, fewer than 30 courses were identified as being studied to complete a module, which is uncommon. Given that there are too few observations to enable module completion to be considered an outcome alongside course completion, these courses were removed from the sample. From a policy perspective, we are most interested in whether pressure from work affects course completion because, if it does, there may be considerable waste of education resources. A complication is in how to treat the significant number of deferrals; moreover, whether deferrals should be treated in the same way as withdrawals (table 6). In this study, a deferral is identified when someone commences study, but gives the reason for stopping as deferral. An individual who enrols, but defers without commencing study, is treated as not being in study. Given this definition, the only difference between deferral and withdrawal in this study is that the former may recommence the same course at a later date, while the latter will not. Therefore, for all intents and purposes, if an individual defers and does not recommence the course, the outcome is the same as withdrawal. In LSAY, we can only identify whether someone recommenced their deferred course in the survey following deferral and we can only identify whether an individual who ends their study with a deferral returns in the following year (13% are identified to recommence in the following year). There is insufficient information in LSAY to identify whether those who return to study after the year following deferral are recommencing their deferred course or are beginning a new course. Without full information on the rates in which deferred students recommence, we treat deferral as a separate category from withdrawal. In terms of completion rates, we observe only small differences between VET and higher education students, but relatively large differences between full-time and part-time students for both groups (table 6). Around 77% of full-time VET students undertaking their first tertiary course are estimated to complete compared with 63% of part-time students. Similarly, around 81% of full-time higher education students undertaking their first tertiary course are estimated to complete compared with 66% for part-time students. Lower completion rates among part-time students may be because they are more likely to combine study and work and are more likely to work longer hours in their jobs (table 3), which may impinge upon their ability to complete study. Alternatively, it may mean that part-time students are less committed to completing their course. To some degree, completion rates in our sample may be overestimated. First, in some waves of LSAY we could not determine whether an individual had been enrolled in between waves of the survey because respondents were only asked whether they are currently in study and whether they had obtained a qualification since the last interview. In these waves, individuals who enrolled and dropped out of their course in between waves would not be in the analysis. Second, the high and non-random respondent attrition rate in LSAY, especially among those from low socioeconomic backgrounds (McKenzie 2002), means that those who are vulnerable to non-completion are underrepresented. These factors apply to both VET and higher education students and it is difficult to judge whether the bias may be more apparent for one group over the other. To the extent that lower socioeconomic students are more highly represented in VET, attrition may lead to a greater overestimate of VET completion. When interpreting results, it is also important to keep in mind that our sample is not representative of all VET students aged 1525 years in their first tertiary course. Module completers (who only enrol in a course subject) are not identified in every wave of LSAY and are excluded from the analysis. As well, we exclude those studying VET as part of an apprenticeship or traineeship. VET students in our sample are mostly higher-end qualifications: diploma students (53%), and certificate level III and IV (29%) (table A2). From table 7, the relationship between hours of work and completion rates appears to be mixed. For full-time higher education and VET students, there appears to be no clear relationship between work and completion. Generally speaking, the relationship for part-time VET and higher education students is similarthose who work are more likely to complete than those who dont, but the rate of completion tends to decline with the hours worked, especially for VET students. The decline in completion rates with the hours of work for part-time students, especially for VET students, is possibly because those who study part-time and work full-time may already have career jobs and are less motivated to complete relative to those who are not working in career jobs (table 3). Table 6 Outcomes of first tertiary course VETHigher educationFull-timePart-timeTotalFull-timePart-timeTotalCount%Count%Count%Count%Count%Count%Completed71377231639447332778132466360179Withdrew15016952624519355966134219Deferred5864111998433111002053312Total921100367100128810040651004901004555100Source: LSAY 1995 and 1998 cohorts. Table 7 Completion rates of first tertiary course by hours of work in the year before ending study VETHigher educationFull-timePart-timeTotalFull-timePart-timeTotalCount%%%%%%Not working79597679507813710.18 hrs7567748684867688.116 hrs787077826681147216.124 hrs75727577677685324.132 hrs77697478647555332.140 hrs785964797176537>40 hrs676565766370241Count9143651279402948745165795Source: LSAY 1995 and 1998 cohorts. The statistics presented in Table 8 do not support this hypothesis and there appears to be no evidence that working in a career job while studying has a negative effect on completion rates. Contrary to expectations, the statistics in table 8 suggest that, for both VET and higher education students, holding a career job may marginally increase the chances of completion relative to those who did not hold a career job while studying. All things being equal, this suggests that students who hold a career job while studying may be more committed to completion because they are more certain of the skill requirements of their future job. However, caution should be exercised because we cannot discount the possibility that the higher completion rates are because those who already have a career job are the best students. This issue will be addressed using multivariate analysis. In summary, from the descriptive statistics presented above, it may be concluded that the impact of working on course outcomes is mixed. In general, there appears to be no clear relationship between working and student outcomes for full-time students. For part-time students, there is some evidence that working does improve the chances of completing, but only up to a point. We find no evidence that working in a career job before ending study reduces the chances of completion. However, these conclusions must be tempered because they do not control for differences in the personal characteristics and situations of various work and study combinations that may affect completion rates. For example, it is possible that students who choose to work full-time and study part-time are less committed to their course than students who study full-time and work part-time. We attempt to control for such differences in the multivariate analysis. Table 8 Found a career job before end of first tertiary course VETHigher educationFull-timePart-timeTotalFull-timePart-timeTotalCount%Count%Count%Count%Count%Count%No career jobCompleted63977144617837429328020563313779Withdrew139176929208203349511638510Deferred506231073738811712245912Total828100236100106410036541003271003981100Found a career job before end of courseCompleted7480876616172345841197346481Withdrew111226203717215159366Deferred8918142612451129187413Total93100131100224100411100163100574100Source: LSAY 1995 and 1998 cohorts. Post-study employment outcomes The second focus of this study is on examining employment outcomes from work and study in the first three years after ceasing study. We limit testing for employment benefits to the first three years after study because sample attrition and response truncation (especially for higher education students) means that extending the survey would increase the rate of missing information. Taking a longer-term view is considered important because it is likely that post-study employment benefits from combining work and study diminish over time. We categorise a job in the last year of study as: 1 one that the individual would like as a career, but is only a casual job 2 one that the individual would like as a career and is an ongoing job (not casual) 3 one that the individual would like as a career and was commenced prior to starting study 4 one that an individual would not like as a career. We separate these types of jobs to try and identify the possible short-term employment benefits of working while studying. It is assumed that the short-term employment outcomes of those who work in a career job they started prior to study (3) and those who found ongoing employment in a career job while studying (2) are not comparable with outcomes of other students who work, because it is not clear whether their post-study employment outcomes are due to work while studying. Therefore, we assume that the employment benefits of combining work and study can only be identified by comparing the post-study employment outcomes of those who worked in a casual career job or a non-career job against the outcomes of those who did not work. We assume that the benefits of working in a casual career job may be linked to the development of job-specific and general skills, whereas the employment benefits of those working in non-career jobs are more related to general skills. In the case of those not working in career jobs, we speculate that employment benefits from generating general skills by working in the last year of study will depend on whether or not they had prior employment experience. We test this hypothesis by interacting years of employment prior to study with the variable identifying those working in a non-career job. As well as information on the type of job, we include controls for the hours worked on average in a week and the years of work experience prior to starting the last spell of study. Descriptive statistics on employment outcomes in the first three years after the last year of study are presented in table 9. For both groups of students, there appear to be benefits in working in the last year of study, with the share of former students in employment, in the short- and longer-term, higher than for those who were not employed in the last year of study. Benefits of combining work and study also seem to be higher for those who were employed in casual career jobs rather than in non-career jobs, especially for higher education students. Table 9 Employment outcomes in the first three years after ending the final spell of study 1st year2nd year3rd yearNot workingEmp. p-t.Emp. f-t.Not workingEmp. p-t.Emp. f-t.Not workingEmp. p-t.Emp. f-t.%%%%%%%%%VET studentsWorked in a casual career job618762217732077Worked in a non-career job734598266692170Didn't work 372638232948212455Count1763235989821049272127380Higher education studentsWorked in a casual career job41482398841185Worked in a non-career job1035555187661579Didn't work 272152141670121573Count5011144214620449822031382971557Note: Emp. P-t. = employed part-time (up to 35 hours per week); Emp. F-t. = employed full-time (more than 35 hours per week). Source: LSAY 1995 and 1998 cohorts. Modelling approach The modelling approach adopted in this study is composed of three parts. The first examines the choice of work and study combinations; the second is academic outcomes from combining work and study; and the third examines the employment benefits from combining work and study. Modelling the choice of study and work combinations To examine the relationship between individual and course characteristics on the choice of study and work combination, we use a multinomial logit (MNL) model. We choose a MNL model because of its flexibilitywe can analyse all work and study combinations together in one model and the estimated coefficients are free to vary across work and study combinations. Alternative models, such as tobit models, place serious restrictions on the choice of work and study combinations. First, such models cannot be used to represent the joint choice of course load and hours of work. Instead, they must assume that the choice of hours in work is conditional on the choice of course load (that is, estimate separate models of hours worked for those who study full-time and part-time), which is not realistic and may bias the results. Second, these models place restrictions on the model coefficients; that is, the estimated sign and significance of coefficients in both the decision to work and decision of hours to work are the same. Under the MNL model, students in higher education and VET are assumed to choose from a discrete number of study and work combinations rather than making sequential decisions: for example, choosing to study part-time and then choosing work hours. Individuals are assumed to assess the benefits and costs (current and future) of each alternative and choose the one that maximises their usefulness or satisfaction. Benefits from working while studying may include increased income while studying and increased likelihood of finding post-study employment. Costs may include foregone time socialising and studying, which may in turn affect the chances of completing study and future income. Potentially there is an infinite number of possible work and study combinations. However, in practice, youth are likely to choose from only a small number of combinations, although the number of choices available may vary between students, depending on individual time constraints. For the purpose of this study, we assume that all young people choose between five alternatives: 1 Full-time study, no work 2 Full-time study, work up to 16 hours per week (approximately two days per week) 3 Full-time study, work more than two days per week 4 Part-time study and work up to 32 hours per week (approximately four days per week) 5 Part-time study and work more than 32 hours per week (full-time employment). These alternatives are chosen because they broadly represent some of the most common combinations and also give us enough observations under each category to estimate robust results. We omit students who are enrolled part-time and who are not working because it is an uncommon combination and there are insufficient observations to allow for estimation of robust results for such a category. Alternative 2 can be considered to include mainly full-time students whose study schedule may not be seriously affected by workthey can comfortably fit their work around their study by working on weekends or outside contact hours. We assume that working more than two days a week while studying full-time may impinge upon study time (alternative 3). For part-time work, which is generally between 25% and 75% of full-time loads, we assume that work for more than four days a week may impinge on study time. A feature of the MNL model, compared with other multi-category models such as the ordered probit, is that each of the study and work alternatives has a separate equation, so that the effect of explanatory variables is allowed to vary across the work and study combinations. However, an important drawback of using an MNL model is that it invokes the independence from irrelevant alternatives (IIA) assumption, which states that the ratio of the probabilities of choosing any two alternatives is independent of the attributes of any other alternative in the choice set (Debreu 1960). In a nutshell, this means that the MNL model assumes that the degree of substitutability between each of the alternatives is the same or that the categories are not ordered. The explanatory variables to be included in this model reflect possible costs and benefits that individuals may receive from various levels of participation. These include: characteristics of the individual, such as their age, gender, size of the city they lived in at the start of the survey characteristics of the individuals course, such as field of education and qualification type characteristics of the individuals parents, such as past employment status, education levels and occupation, whether they help pay tuition fees, which may reflect the socioeconomic status of parents parental wealth and income, proxied by their post-school qualifications, employment while students are in school and using the ANU3 scale of the status of both parents occupations whether a student receives income support through Youth Allowance to study academic ability, proxied by their university entry scores living arrangements, such as whether or not they are still living at home other commitments, such as whether or not they have children. A point of note is that we do not include the amount of Youth Allowance in the model because it is highly likely to be endogenous; that is, simultaneously determined with the choice of work and study combination. A limitation of LSAY is that we cannot identify an individuals place of residence while they were studying, only their location when they joined the survey in Year 9. Therefore, we cannot examine specific impacts on students from rural areas, but we can examine more generally the impacts of students living away from home. A limitation of using the multinomial logit approach is that the same variables must be included for each of the equations. This means that we cannot examine how employer characteristics may affect the choice of hours worked because such information is not observed for those individuals who choose not to work. For some variables of interest, such as field of study, while there are observations for each of the work and study combinations, there are insufficient numbers of observations for some categories, which makes estimation problematic. Such variables were not included in the analysis. Modelling education outcomes In this study, we analyse the impacts of combining work and study by examining the impacts of average hours of work during study on three student outcomeswithdrawal, deferment and completion. Given that these three outcomes are orderedthat is, they represent three outcomes along a scale of academic achievementwe estimate the impacts of combining work and study using an ordered probit model. We estimate models for part-time students and full-time students separately to try to adjust for the effects that different academic workloads may have on students ability to manage work and study. All else being equal, working three days a week might have negligible impacts on the academic outcomes of part-time students, but the impacts of such hours are likely to be more negative for full-time students. Further, we split the two groups to control for unobserved differences between part- and full-time students, which may also affect completion rates. These may include differences in the perceived employment benefits from completing, differences in individuals circumstances (such as whether they have other pressing commitments; for example, family commitments) and differences in personality traits, such as motivation and persistence. To further control for the academic demands of a course, we include information on the field of education and type of course. The ordered probit model, an extension of the binary probit model, is based on the assumption that there exists an underlying index functionin this report, an underlying index of academic performance that can be mapped to the three academic outcomes. We assume that the underlying index depends upon a range of individual and employment characteristics. Individual characteristics to be included in the model include past educational performance, proxied by the Interstate Transfer Index, which is a standardised index of university entry scores. Year 9 numeracy and literacy scores as well as a combined index were trialled in the model instead of using the Interstate Transfer Index, but none was significant. Other personal characteristics include employment history, socioeconomic status and employment status of parents, age, gender, marital status, presence of dependent children, whether or not the student had a break from study before enrolling in their first tertiary course and state of residence. Employment information includes work history, tenure of employment, occupation and hours of work. Because part-time and full-time students have different patterns of work hours, we include different levels of commitment in each. For part-time students, hours of work are categorised as part-time (working up to 32 hours a week) and full-time (more than 32 hours of work a week). For full-time students, hours of work are none, up to eight hours per week (approximately one day), 8.1 to 16 hours per week (between one and two days), 16.1 to 24 hours per week (between two and three days) and more than 24 hours per week (more than three days per week). These levels were chosen because they are common combinations, which are the same combinations used in the model of work and study choices. There are also enough observations under each category to give robust results. As for the modelling of work and study choices, we omit those who study part-time and do not work. There are few of these individuals and they are more likely to be working part-time due to circumstances that are unrelated to the circumstances of other part-time students; for example, because of possible health conditions. As well as estimating an ordered probit model of student outcomes, we estimate a bivariate ordered probit model of hours worked and student outcomes on a pooled sample of full-time students to test for self-selection on unobservables that may bias the results. As discussed above, a possible issue when modelling the effect of working while studying on student outcomes is controlling for possible self-selection on unobserved factors, such as student performance and attachment to the course, which may be correlated with both hours of work and outcomes. A bivariate ordered probit model involves estimating an ordered probit model of hours worked and an ordered probit model of student outcomes jointly, with self-selection bias controlled for by allowing the error terms from these two equations to be correlated. To identify this model, we include a number of identification restrictions that are assumed to affect hours of work, but not student outcomes, including population of the place of residence (but not where the education occurs) and whether each parent is a manager. It is assumed that having a parent as a manager is an important pathway for employment while in study, but after controlling for socio-demographics of parents, it does not affect outcomes from study. Modelling employment benefits from combining work andstudy As discussed above, we examine not only short-run employment benefits from combining work and study, but also benefits up to three years after completion. Other potential labour market benefits of work and study, such as job satisfaction and wage progression, are not examined here. To estimate the longer-term benefits of employment, we estimate a random effects ordered panel probit model, using the first three waves of data after the last course undertaken in the LSAY dataset. The three outcomes modelled are: not employed (coded 0), employed part-time (coded 1) and employed full-time (coded 2). We use an ordered model of employment outcomes rather than a binary model of employment (employed or not) because it gives us more information about the actual outcomes from combining work and study. Because a binary employment model does not distinguish between full-time and part-time study, it is likely that it would tend to overestimate the employment benefits from combining work and study: it is possible that, while looking for work, those who completed study may continue to work in their non-career job. The variables included in the model are those that relate to the characteristics of employment, such as hours worked, type of job and whether the individual had prior work experience. We also include factors such as marital status and the presence of children, and socioeconomic background, which may affect both the choice of working hours while studying and post-study employment outcomes. To test how the employment outcomes from combining work and study vary through time, we interact a variable identifying years after study with the above job types while studying. Because we are only interested in the outcomes for a given year and not the path that individuals follow from year to year, we do not estimate a dynamic model (which includes a lagged dependent variable as an explanatory variable). We estimate the employment model of post-study outcomes for VET and higher education students, which allows us to test for differences between outcomes for VET and higher education students. Results Results from the multivariate models estimated in this report are average marginal effects and represent the estimated average (across the sample) percentage-point change in the probability of the outcome or dependent variable for a one-unit change in each of the explanatory variables, independent of the effects of all other explanatory variables in the model. For categorical variables, the marginal effects represent the percentage-point change in the probability of a given outcome, relative to the reference category that is omitted. In most cases, we report a t-statistic for each marginal effect. A t-statistic indicates the precision of the estimate and can be interpreted as a test of the confidence with which we can state that a marginal effect is different from zero. Although it is customary to check if a t-statistic is higher than 1.96, which is the value representing a 5% level of significance, it is good to remember that the higher the t-statistic, the more precise is our estimate. In other cases we refrain from reporting the t-statistic to save space, but instead just report the level of significance as either 5% or 10%. Choice of work and study combinations Results from the multinomial logit model of work and study combinations for first-time tertiary students are presented in table 10. In this context, the marginal effects represent the percentage-point change in the probability of choosing a specified work-study combination, given a one-unit change in an explanatory variable, all else being equal. For example, from table 10 we can say that those who are studying a VET qualification are eight percentage points more likely to study full-time with no employment than higher education students (reference category). Because these combinations represent an exhaustive set (after excluding those who work part-time and do not work), the sum of the marginal effects across the five combinations is zero. In other words, if increasing the level of an independent variable increases the likelihood of choosing four of the combinations, it must reduce the probability of choosing the remaining combination. A clear result from table 10 is that VET and higher education students prefer different work and study combinations. Generally speaking, the results show that young people undertaking VET are more likely than higher education students to study full-time without working, which may be because VET courses are typically shorter and hence there is a lesser need to supplement their income. However, when combining work and study, VET students are more likely to study part-time than higher education students, who are more likely study full-time. The higher rate of VET students combining part-time study and work rather than full-time study and work may be because of the more flexible ways in which VET students may undertake training, which gives them greater scope for part-time study and work combinations. Table 10 Marginal effects for choice of work and study combination for first tertiary course Full-time studentPart-time studentNo workUp to 2 days p.w.More than 2 days p.w.Up to 4 days p.w.More than 4 days p.w.m.e.t-statm.e.t-statm.e.t-statm.e.t-statm.e.t-statConstant0.091.45-0.32-4.33**0.192.77**0.020.700.031.441995 cohort-0.02-1.390.021.610.021.50-0.01-2.56**-0.01-3.21**Female-0.08-6.46**0.074.78**0.021.260.000.53-0.01-2.96**Married or de facto-0.04-1.43-0.02-0.520.020.630.021.430.023.73**Place of birth (ref: born in Australia)Born in other English speaking country-0.03-0.720.051.19-0.01-0.330.000.27-0.01-1.11Born in non-English speaking country0.105.17**0.010.46-0.08-3.25**-0.01-1.35-0.02-2.41**Age (ref: less than 18 years old)1820 years old-0.04-1.490.030.800.020.580.000.06-0.01-1.042123 years old-0.07-2.96**0.020.650.00-0.020.032.16**0.033.58**Older than 230.010.18-0.09-1.46-0.06-0.990.085.00**0.065.22**University entry score (ITI x 0.1)10.000.370.035.00**-0.02-3.49**-0.01-3.41**0.00-3.66**University entry score missing0.071.97**0.142.98**-0.16-3.67**-0.03-1.66-0.03-2.96**Live with parents-0.10-5.09**0.166.75**-0.03-1.75*-0.01-1.33-0.02-3.70**State of residence in initial survey (ref: NSW)Victoria-0.02-1.150.031.670.021.06-0.01-1.48-0.02-3.83**Queensland0.00-0.03-0.06-2.73**0.052.68**0.000.360.000.71Western Australia-0.01-0.680.00-0.060.010.500.000.220.000.35South Australia0.031.54-0.04-1.70*0.00-0.030.011.140.000.11Tasmania0.072.36**0.010.15-0.05-1.33-0.02-1.27-0.01-0.60ACT and NT0.010.40-0.03-0.840.010.420.000.210.000.15Population of place of residence (ref: More than 100 000)100099000.052.89**-0.01-0.38-0.04-2.28**0.00-0.150.000.87Less than 10000.063.68**-0.01-0.68-0.05-2.40**0.00-0.280.000.31Enrolled in a VET course0.084.23**-0.07-2.91**-0.08-3.88**0.033.75**0.047.24**Receives Youth Allowance0.136.55**0.228.49**-0.15-6.17**-0.10-7.60**-0.11-9.03**Father's highest post-school qualification (ref: none)Certificate-0.05-3.09**0.031.580.020.920.000.400.000.42Higher education0.00-0.150.041.71*-0.03-1.66-0.01-0.800.011.13Unknown-0.03-0.970.041.26-0.03-0.950.010.930.000.42Father is employed-0.01-0.560.000.110.020.54-0.01-0.880.000.46Father is a manager0.00-0.01-0.04-2.12**0.031.82*0.000.620.000.95Father's ANU3 occupation score0.00-0.530.000.920.00-0.080.000.620.00-2.46**Mother's highest post-school qualification (ref: none)Certificate0.021.03-0.03-1.290.010.450.000.020.00-0.08Higher education0.000.07-0.01-0.740.021.090.000.36-0.01-1.85*Unknown0.041.58-0.02-0.66-0.01-0.180.00-0.19-0.01-1.51Mother is employed-0.04-2.23**0.010.590.041.92*-0.01-1.530.00-0.08Mother is a manager0.010.35-0.06-1.76*0.051.71*-0.01-0.790.011.36Mother's ANU3 occupation score0.000.290.000.430.00-1.280.000.990.000.54Youth Allowance x Live with parents0.072.63**-0.07-2.10**-0.05-1.72*0.031.670.031.47Sample size5660Notes: ** Significant at 5%; *Significant at 10%. 1 Missing university entry scores are replaced with combined Year 9 mathematics and numeracy scores scaled to between 0 and 1. Irrespective of the type of study, those with a higher university entry scoreobviously a measure of academic abilityhave a preference for working full-time up to two days per week. For example, all else being equal, those who have an Interstate Transfer Index of 75% are estimated to be 15 percentage points more likely to work full-time up to two days work per week than someone with an index of 25% and approximately ten percentage points less likely to be employed part-time. These results suggest that part-time students may be less academically inclined than their full-time counterparts. However, we should point out that the distribution of entry scores between the two groups is quite different and it is difficult to conclude whether the effect of past academic achievement is consistent for VET and higher education students. Age is a strong predictor of the likelihood of part-time enrolment, with those who re-engage in education at a later age much more likely to choose the part-time study option. Relative to those who re-engage before 18, those who re-engage at age 23 or older are around eight percentage points more likely to be enrolled part-time and working up to four hours per week and six percentage points more likely to be enrolled part-time and working more than four hours. The greater popularity of part-time work among older students is probably because they already have employment and are less inclined to forego income and their standard of living for full-time study. An important finding from the results is that the choice of work and study combination does not appear to be strongly linked to the socioeconomic status of parents. All else being equal, if working while studying was closely linked to the familys financial resources, we may expect that those from lower socioeconomic backgrounds would be less likely to be enrolled full-time and not working. However, results from table 10 show that parents working status, occupation status (measured by the ANU3 scores) and tertiary qualifications are not strongly related to choice of work and study combination. This result is consistent with findings from analysis reported by the House of Representatives Standing Committee on Education and Training (2009), which found that rather than work for financial need, school students tend to work for financial independence and to support their lifestyle. Results do suggest, however, that students whose parents are managers are more likely to be studying full-time and work more than two days per week. Cultural factors appear to play a role in the choice of work and study combination, with children born in non-English speaking countries around ten percentage points more likely to be enrolled full-time without work compared with Australian-born students. After controlling for the socioeconomic status of parents, students who receive Youth Allowance are much more likely to be in full-time study and spend less time in work. However, we find that the effect of receiving Youth Allowance on the probability of working while in full-time study is greater for those who live at home than for those who live away because they are likely to have fewer financial needs. For those living at home, receiving Youth Allowance is estimated to increase the chances of being in full-time study and not working by 20 percentage points compared with 13 percentage points for those living away from home. While this may suggest that Youth Allowance reduces the incentive to work, to some extent the relationship may work the other way, those who prefer to work more than two days a week are not eligible for Youth Allowance. Finally, the choice of work and study combination is also affected by local employment opportunities. Those from rural (population of more than 1000 but fewer than 100000) and remote areas (population fewer than 1000) are five percentage points and six percentage points more likely tobe studying full-time without work than students from urban areas (population over 100000). Student outcomes To understand how combining work and study affects student outcomes, we estimate univariate ordered probit models of the first tertiary course outcomes for part-time and full-time students separately. If we had pooled the data for the full-time and part-time students, it would be difficult to estimate the effect of work because there are likely to be considerable differences in the academic requirements between part and full-time students. The results are presented as marginal effects, or the change in the probability of each of the three outcomes (withdraw, defer and complete) for a one-unit change in an explanatory variable. Because the student outcomes are exhaustive, the sum of the marginal effects for a given explanatory variable is zero. From the highly significant estimated threshold parameters ( s) in each of the sets of results in table 11, it can be concluded that student outcomes are clearly ordered. That is, they represent different points along a scale of academic achievement, which means that an ordered model is appropriate. When interpreting these results, it is important to keep in mind the small sample sizes for part-time students, which means that the estimated results for part-time students are less precise (and thus few significant results for part-time students). In cases where the marginal effects are large, but insignificant, there may be reason to investigate the result further. Table 11 Marginal effects of student outcomes, all first-time tertiary courses Full-time studyPart-time studywithdrewdeferredcompletedwithdrewdeferredcompleted1995 cohort-0.003-0.0030.006-0.041**-0.030**0.070**Has children0.191***0.094***-0.286***-0.012-0.0090.020Female-0.025***-0.022***0.046***-0.028-0.0200.048Live with parents-0.004-0.0040.008-0.018-0.0130.031Married or de facto-0.021-0.0210.042-0.042-0.0350.078Place of birth (ref: born in Australia)Born in other English speaking country-0.017*-0.017*0.034*-0.083**-0.084**0.167**Born in non-English speaking country0.042**0.033**-0.075**-0.022-0.0180.040ANU occupation socioeconomic of father (110)-0.002-0.0020.003-0.002-0.0020.004ANU occupation socioeconomic of mother (110)0.0000.0000.0000.0060.004-0.010Age (ref: less than 18)18200.023*0.02*-0.043*0.096**0.061**-0.157**2123-0.073***-0.064***0.137***-0.067-0.0480.115Older than 23-0.054***-0.065***0.119***-0.019-0.0150.034State of residence in initial survey (ref: NSW)Victoria 0.0110.01-0.0210.0040.003-0.007Queensland 0.0120.01-0.0220.0540.035-0.088Western Australia -0.005-0.0050.010.0280.019-0.047South Australia 0.0080.007-0.0150.0260.018-0.044Tasmania -0.003-0.0030.0050.0200.014-0.033ACT and NT0.041***0.032***-0.073***0.103**0.057**-0.159**University entry score (ITI x 0.1)1-0.192***-0.172***0.364***-0.204***-0.120***0.325***University entry score missing0.0100.010-0.0200.0290.0160.045Enrolled in a VET course-0.048***-0.05***0.098***-0.059**-0.045**0.104**Field of study (ref: education)Unknown-0.049**-0.057**0.107**-0.136-0.1830.319Nat. & phys. sciences0.0250.021-0.046-0.041-0.0350.077Information technology0.041**0.032**-0.073**-0.013-0.0100.023Engineering0.054***0.041***-0.094***-0.043-0.0370.080Architecture & building0.0440.034-0.0780.0520.032-0.084Agric., environ. & related0.0020.001-0.003-0.069-0.0670.135Heath-0.014-0.0130.027-0.016-0.0120.028Management & commerce0.0060.005-0.011-0.022-0.0160.038Society & culture0.034**0.029**-0.063**0.0240.016-0.040Arts0.029*0.024*-0.054*-0.014-0.0110.025Other fields of study0.0090.008-0.016-0.033-0.0270.060Completed school-0.014-0.0120.027-0.045-0.0290.074Started study more than a year after school0.027**0.022**-0.049**0.0270.019-0.046Start time missing-0.028**-0.029**0.057**0.0510.032-0.083Employment characteristics in last year of studyJob would like as a career (ref: not a career job)Career job found before study 0.0440.034-0.0780.0320.021-0.053Career job found while studying-0.019**-0.019**0.038**-0.057***-0.045***0.102***Years worked in current job-0.005**-0.005**0.010**-0.016**-0.012**0.029**Years of past employment -0.013***-0.012***0.025***-0.007-0.0050.012Average hours worked during courseFull-time study (ref: no work while studying)Up to 8 hours0.0000.000-0.001---8.116 hours0.024**0.021**-0.046**---16.1 24 hours0.043***0.035***-0.079***---More than 24 hours0.084***0.059***-0.143***---Part-time study (ref: work up to 32 hours per week)Work more than 32 hours0.070***0.047***-0.117***0.51**0.628***Sample size4931853Notes: ***Significant at 1%; **Significant at 5%; *Significant at 10%. 1 Missing university entry scores are replaced by combined Year 9 mathematics and numeracy scores that are scaled to between 0 and 1. A key result from table 11 is that, after controlling for a range of differences between VET and higher education studentsincluding academic ability (measured by university entry score or combined Year 9 numeracy and literacy scores if entry scores are missing), course type and field of studyfull-time and part-time VET students are both around ten percentage points more likely to complete than their higher education counterparts. The higher completion rates among VET students may be due to a number of reasons, including the shorter duration of the courses, differences in academic demands and differences in the flexibility and modes of course delivery. In general, the modularised nature of VET means that courses can be better tailored to individual training needs and are delivered in a greater range of modes, especially off-campus delivery modes. For full-time students, we find that those who work beyond eight hours (roughly one working day) a week on average over the course of their study are less likely to complete than those who do not work at all while studying. Full-time students who work for between eight and 16 hours a week (roughly two working days) on average over the course of their study are estimated to be around five percentage points less likely to complete compared with those who do not work at all. At the extreme, full-time students who work more than 24 hours a week (roughly more than three days) are estimated to be around 14 percentage points less likely to complete compared with those who do not work at all. For part-time students, due to the small number of observations, the only comparison is between those who work fewer than 32 hours per week (part-time workers) and those who work more than 32 hours (full-time workers). We estimate that part-time students who work more than 32 hours per week are around 12 percentage points less likely to complete than those who work less than 32 hours per week. From models with interaction effects (table 12) we find that, after controlling for differences between VET and higher education students, there is no evidence that working while studying full-time affects VET and higher education students differently. An exception is that VET students who work up to eight hours a week are estimated to complete at an 11-percentage point lower rate than higher education students. From analysis of the data, it appears that this result is driven by lower completion rates among diploma/advanced diploma health students who report working up to eight hours a week and may reflect the effect of work placements that are often associated with these courses. However, there is insufficient information in LSAY to identify whether the work is part of a placement. From the positive and significant interaction effect in table 12, we can conclude that part-time VET students who work full-time (work more than 32 hours per week on average over their course) are around 15 percentage points more likely to complete than part-time higher education students who work full-time. To complete a qualification part-time while working full-time requires considerable effort and application and the longer duration of higher education courses may make the required commitment more taxing. The relatively long commitment required to obtain a higher education qualification part-time may also mean that employers may be less likely to support full-time employees who choose this education pathway. Table 12 Key interactions between VET and hours of work in first tertiary courses Full-timePart-timewithdrewdeferredcompletedwithdrewdeferredcompletedStudying for a VET qualification-0.054-0.0580.112***-0.025-0.0190.044Average hours worked during courseFull-time study (ref: no work while studying)Up to 8 hours-0.009-0.0080.016---8.116 hours0.0210.019-0.039*---16.124 hours0.0380.031-0.069***---More than 24 hours0.0970.066-0.163***---Part-time study (ref: work up to 32 hours per week)Work more than 32 hours0.1280.079-0.206***Study type and work interactionsVET x FT Work up to 8 hours0.0640.047-0.111**---VET x FT Work 8.116 hours0.0150.013-0.029---VET x FT Work 16.124 hours0.0280.023-0.050---VET x FT Work more than 24 hours-0.023-0.0230.046---VET x PT Work more than 32 hours----0.080-0.0720.152***Notes: ***Significant at 1%; **Significant at 5%; *Significant at 10%. It is estimated that the effect of work depends on the nature of the job. Those who find a career job while studying are more likely to complete than those who work in a non-career job. Those who find a career job while studying are mostly finding work in professional jobs, especially in information technology, engineering, and architecture and building. Because professional jobs tend to require the attainment of a qualification for post-study employment, students who find work in these areas may have an added incentive to complete. Further, their employers may provide them with support, such as favourable work arrangements, to help them to attain their qualification. Although we control for academic ability in the form of university entry scores (or combined Year 9 numeracy and literacy scores if entry scores are missing), we cannot rule out the possibility that those who find career jobs while studying are better-performing students, who are more likely to complete. In contrast, we find evidence that those who continue to work in a career job that was held prior to study (including those who change to another job while studying) are less likely to complete than those who worked in a non-career job. Full-time and part-time students who continue working in a career job they had prior to study are estimated to be around eight and five percentage points respectively less likely to complete than those working in a non-career job. There may be a number of factors combining to produce this result. First, those who continue working in a career job while studying may only intend to complete the parts of the course relevant to their job. Second, those working in career jobs may not have the same incentives to complete as those who do not work in a career job. However, both of these estimated effects are just outside the 10% level of significance,most likely because of the small number of students aged 1524 with career jobs found prior to study. A consistent result for part- and full-time students is that the longer an individual has been in the job, the greater the chance of completion. A possible explanation is that the more established an individual is in the job, the more support they get from their employer in the form of more flexible working hours or possibly time off work, in the case of full-time employees. For full-time students, the more years of employment experience, the greater the chance of completion. There may be two possible reasons for this. On the one hand, employment history, including history in the same job, may help develop a range of soft skills such as time management, commitment to completing a task, communication and interpersonal skills, as well as confidence, which may help academic performance. On the other hand, the relationship may not be causal, but instead related to uncontrolled differences in the characteristics of students who have and have not a history of working; for example, differences in motivation. Using interaction terms, we found no evidence that the effect of work and study combinations while studying was related to employment history. To test the robustness of the effect of average hours in work during study, we estimated a range of alternative models. First, to test for evidence of self-selection bias, we extended the univariate ordered probit model (results discussed above) to a bivariate ordered probit model. Under this model, we control for the effect of possible unobserved factors that affect both education outcomes and hours of work, such as attachment to course. Results from this model using different exclusion restrictions, including region and parents socioeconomic status, failed to show any significant evidence of self-selection on unobservables. As a further test of the adequacy of observed controls, we omitted hours of work and re-estimated the model. If there existed unobserved factors that were correlated with both hours of work and completion, then we would expect omitting hours of work would significantly affect the results, especially for factors that are correlated with hours of work, such as past academic performance and type and field of study. However, omitting hours of work made little difference to the results, including on the estimated coefficients for past academic performance and type and field of study. This further supports the results from the bivariate probit model. We also find that treating the outcome variable as a three-category outcome rather than a binary outcome (study ended with a qualification or not) has very little bearing on the results, with estimated marginal effects of completion for a binary probit model comparable with those reported in table 11. Finally, we tested how choosing the hours worked in the last period of study, instead of average hours worked while in study, would affect the results for full-time students in table 11. We found that using the same categories of work hours (none, up to eight hours, 8.116 hours, 16.124 hours and more than 24 hours) but from the last period of study considerably increases the estimated negative relationship between hours of work and the chances of completion, especially for the lower hour categories. To the extent that students who studied for more than one period and who commenced work after their first year of study may have already had a reduced chance of completion before starting work, using work in the last period of study may overestimate the impacts of combining work and study on course completion. Employment outcomes For the purposes of this section, we present predicted post-study probabilities of full-time employment for those who combined study and those who did not (table 13) (complete results from the employment model are presented in table A3 in the appendix). The predicted full-time employment probabilities are calculated for an individual with average characteristics, except for their past employment status. To calculate the employment probability for those who worked while studying, it is assumed that between 8.1 and 16 hours are worked per week. When estimating predicted probabilities for those who had work experience prior to studying, we assume that they had only one year of prior employment experience. The standard errors (and hence the level of significance) of the marginal effects are estimated using the delta method. Table 13 Predicted probabilities of full-time employment in the first three years after ending study VETHigher educationDifference in ME1st year2nd year3rd year1st year2nd year3rd year1st year2nd year3rd yearHad not had prior work experienceDidn't work in last year of study (reference case)0.230.720.820.110.860.92---Worked in a non-career job in last year of study0.820.990.990.270.950.99---Marginal effect of working while studying0.58***0.26***0.18***0.16***0.09***0.07**0.42***0.17***0.11***Had prior work experienceDidn't work in last year of study (reference case)0.180.780.860.080.900.94---Worked in a non-career job in last year of study0.830.990.990.320.961.00---Worked in a casual career job in the last year of study0.860.880.880.810.970.95---Marginal effect of working in non-career job while studying0.65***0.20***0.13***0.25***0.06**0.05**0.40***0.14***0.08***Marginal effect of working in a casual career job while studying0.68***0.10**0.010.74***0.07**0.00-0.06**0.03**0.01Effect of prior work experience for those who worked in a non-career job0.010.000.000.05**0.010.01-0.04**-0.00-0.00Notes: ***Significant at 1%; **Significant at 5%; *Significant at 10%; ME = marginal effects. Predictions are made for an individual with average characteristics except for their work history and work while studying. For individuals who worked while studying, predictions are made assuming that they worked between 8.1 and 16 hours per week. The significance of the marginal effects is calculated by deriving standard errors of the estimates using the delta method. From table 13, we can conclude that those who work in the last year of study are much more likely to be in full-time employment in the first year after ceasing study than those who did not work while studying. In the first year out from study, on average, VET students who worked in a casual career job in their last year of study are estimated to be 68 percentage points more likely to be in full-time employment in their first year out compared with those who did not work in their last year. Similarly, higher education students who worked in casual career jobs are estimated to be 74percentage points more likely to be in full-time employment in their first year out compared with those who did not work in their last year. The high initial full-time employment benefits of working in a casual career job are likely to be because students continue to work in these jobs in the year after studying, which may not be the case for students who work in jobs that are not career jobs, especially higher education students. The initial full-time employment benefits for higher education students who worked in a non-career job in their last year of study is only around a third of that for those who worked in a casual career job (25-percentage point increase in the full-time employment probability compared with 74-percentage point increase). For VET students, there is only a three-percentage point lower full-time employment benefit from working in a non-career job instead of a casual career job. The reason for the difference in the initial benefits from working in non-career jobs between VET and higher education students is difficult to pinpoint. A possible explanation is that employers of higher education graduates may not value general skills to the same degree as employers of VET graduates, possibly because general skills may not be as important in performing job tasks. Another explanation is that employers of higher education graduates value general skills just as highly, but consider course completion as a better measure of these than working while studying. Finally, although they report that they do not want the job as a career, by comparison with higher education students, VET students may be more likely to derive job-specific skills that are recognised by employers from a non-career job. However, we cannot rule out the possibility that there are unobserved factors, such as personal traits, that affect both their decision to work in the last year of study and also their post-study employment prospects. To the extent that such factors are present, the post-study employment benefits of working while studying will be overestimated to some degree. An important finding is that, although the employment benefits of working in a non-career job decline over time, they are still significant up to three years after ending study. Compared with those who did not work in the last year of study, VET students who worked in a non-career job are 13 percentage points more likely to be in full-time employment and higher education students are five percentage points more likely. The longer-term full-time employment benefit of working in non-career jobs in the last year of study suggests that, not only does combining work and study help students find full-time employment, but it also helps them stay in full-time employment, by sorting them into suitable jobs and developing their skills. In contrast, despite having higher initial full-time employment benefits, the employment benefits of working in a casual career job in the last year of study diminish after the second year for both VET and higher education students. The rapid tapering-off of full-time employment benefits for those who worked in casual career jobs in the last year of study may be because they do not move to ongoing employment and hence are more likely to lose their jobs. Finally, we find that work experience prior to the last year of study makes little difference to post-study employment chances for those who worked in a casual non-career job in their last year of study. The estimated initial post-study employment probability of higher education students working in casual non-career jobs in their last year of study is estimated to be five percentage points higher if they have a record of prior work experience. However, this effect disappears after the first year out from study. For VET students, prior work experience makes no significant difference to employment. Conclusions The objective of this study was to examine the factors affecting tertiary student choice of work and study combinations and the effect of such choices on student and study outcomes. A particular focus was on trying to identify differences in the choices and outcomes between VET and higher education students. Such information is important in being able to design well-targeted measures for supporting students while in study. Consistent with the findings of the House of Representatives Standing Committee on Education and Training (2009), which focused on the implications of working while at school, we find little evidence that the financial and human resources of parents affect the work and study combinations of students. However, we note that the analysis is on tertiary courses that were mostly undertaken prior to 2006 and we cannot rule out the possibility that rising costs of living may make this result redundant. In particular, if the availability and amount of Youth Allowance has not risen to take into account the rise in the costs of study (including living costs), then the estimated effect of Youth Allowance in reducing the need and hours worked by youth may be overstated in this report. Estimated outcomes in this report from combining work and study highlight both the costs and benefits of this practice. On the cost side, depending on the hours worked, results show that students who work while studying may be less likely to complete than those who do not. For full-time students, the negative impacts of work on completion kicks in beyond eight hours of work (roughly a day) per week on average. For part-time students, full-time work is estimated to significantly reduce the chances of completion relative to part-time work. We find no evidence that these results are biased by self-selection into hours of workwhere the choice of hours of work and completion are affected by unobservable variables, such as attachment to course. The estimated negative education outcomes are robust to a range of assumptions. In terms of benefits, results presented in this study suggest that working while studying is more than a signal of motivation that assists students to find work; it also provides them with skills that help maintain them in employment. We find that working in the last year of study improves the chances of students finding full-time employment, up to three years after ceasing study. For VET students, unlike for higher education students, the post-study employment benefits are not strongly linked to having worked in a job that may be directly relevant to their career. Possible explanations for this include employers of higher education graduates not valuing general skills to the same degree as employers of VET graduates, possibly because general skills may not be as important in performing job tasks. Or, although they do not want the job as a career, VET students, by comparison with higher education students, may be more likely to derive job-specific skills that are recognised by employers. For policy-makers, results from this study demonstrate the possible positive and negative affects from measures of income support. On the one hand, income support may reduce the need to work, which will help completion and post-study employment outcomes. On the other hand, income support, by reducing hours of work, may limit opportunities for developing general and job-specific labour market skills. As well as being cognisant of these potential trade-offs, policy-makers should attempt to reduce any influence of work hours on the likelihood of study completion; for example, by enticing providers to offer more flexible modes of education. Another possibility, which arises from findings from this study, is to assist students in finding work that better matches their intended post-study career. Analysis presented in this study shows that those who find a job they would like as a career are more likely to complete study than those who work in jobs that they do not want as a career. Working in a career job may provide added incentives for students to complete because this experience may help them to better understand the importance of course material in performing jobs that they would like as a career and/or give them an extra incentive to complete to obtain ongoing employment. One way to achieve such an outcome would be for governments to support campus brokerage services to link education providers to suitable employers. References Abhayaratna, J, Andrews, L, Nuch, H & Podbury, T 2008, Part time employment: the Australian experience, Staff working paper, Productivity Commission, Canberra. ABS (Australian Bureau of Statistics) 1997, Australian Standard Classification of Occupations (ASCO), Second edition, ABS cat.no.1220.0, Australian Government Publishing Service, Canberra. 2006, Australian and New Zealand Standard Classification of Occupations (ANZSCO), First edition, ABS cat.no.1220.0, Australian Government Publishing Service, Canberra. 2008, Yearbook of Australia, 2008, ABS cat.no.1301.0, ABS, Canberra. Carr, R, Wright, J & Brody, C 1996, Effects of high school work experience a decade later: evidence from the national longitudinal study, Sociology of Education, vol.69, January, pp.6681. Debreu, G 1960, Review of R Luce individual choice behaviour, American Economic Review, vol.50, pp.1868. Dwyer, P, Harwood, A, Costin, G, Landy, M & Wyn, J 1999, Combined study in work paths in VET: policy implications and analysis, ϲʿ¼, Adelaide. House of Representatives Standing Committee on Education and Training 2009, Adolescent overload? Report of the inquiry into combining school and work: supporting successful youth transitions, Parliament, Canberra. Marks, G, Fleming, N, Long, M & McMillan, J 2000, Patterns of participating in Year 12 and higher education in Australia, LSAY research report no.17, Australian Council for Educational Research, Melbourne. McInnis, C & Hartley, R 2006, Managing study and work: the impact of full-time study and paid work on the undergraduate experience in Australian universities, Evaluations and Investigations Programme report no.02/06, Department of Education, Science and Training, Canberra. McKenzie, P 2002, The Year 9 class of 95 in 99, activities and experiences, Australian Council for Education and Research, Melbourne. McMillan, J & Jones, F 2000, ANU3_2 scale: a revised occupational status scale for Australia, Journal of Sociology, vol.36, no.1, pp.6472. McMillan, J, Rothman, S & Wernert, N 2005, Non-apprenticeship VET courses: participation, persistence and subsequent pathways, Australian Council for Educational Research, Melbourne. Meyer, R & Wise, D 1982, High school preparation and early labor force experience, in The youth labor market problem: its nature, causes, eds R Freeman and D Wise, University of Chicago Press, Chicago. ϲʿ¼ (ϲʿ¼) 2009, Australian vocational education and training statistics: students and courses 2008, ϲʿ¼, Adelaide. Riggert, S, Boyle, M, Petrosko, J, Ash, D & Rude-Parkins, C 2006, Student employment and higher education: empiricism and contradiction, Review of Educational Research, vol.76, no.1, pp.6392. Robinson, L 1996, School students and part-time work, LSAY research report no.2, Australian Council for Educational Research, Melbourne. 1999, The effects of part-time work on school students, LSAY research report no.9, Australian Council for Educational Research, Melbourne. Ruhm, C 1997, Is high school employment consumption or investment?, Journal of Labor Economics, vol.14, no.4, pp.73576. Stanwick, J 2005, Australian Qualifications Framework lower-level qualifications: pathways to where for young people, ϲʿ¼, Adelaide. Sweet, R 2008, The job ready certificate, a discussion paper for the Department of Education, Employment and Workplace Relations, Canberra. Tinto, V 1993, Leaving college: rethinking the causes and cures of student attrition, University of Chicago Press, Chicago. Vickers, M, Lamb, S & Hinkley, J 2003, Student workers in high school and beyond: the effects of part-time employment on participation in education, work and training, LSAY youth research report, Australian Council for Educational Research, Melbourne. Appendix Table A1 Employment rates and hours of work during higher education students first tertiary course Length of studyspellYear of spellNumber countEmployed %Avg. hours worked per weekOne year15237419Two years1967691729727818Three years113536715213587617313527919Four years110586415210697416310677918410667919Five years1621651626297217363079194628822056288019Source: LSAY 1995 and 1998 cohorts. Table A2 Level of VET qualification studied for those aged 1525 whose first tertiary course is a VET course AQF qualification%Certificate III11Certificate IIIIV29Certificate unknown17Diploma and advanced diploma53Count (N)1313Note: 1 Non-response or respondent could not recall the level of the certificate Table A3 Coefficients for the ordered probit model of post-study employment outcomes coeff.t-stat.Constant|2.014.46**1995 cohort-0.17-1.57Time period (ref: 199699)200001-0.42-1.53200003-0.52-2.75***200005-0.49-3.28***2000070.201.47Has children-2.78-16.08***Has a disability-0.85-3.37***Female-0.11-1.61Married or de facto0.010.14State of residence in initial survey (ref: NSW)Victoria-0.10-1.03Queensland-0.28-2.53***Western Australia-0.04-0.36South Australia-0.18-1.49Tasmania-0.31-1.83*ACT and NT0.00-0.01Place of residence (ref: city/town with more than 100 000 residents)Less than 100 000 residents-0.18-2.01**Less than 10 000 residents-0.17-1.76*Numeracy and literacy score (110)0.142.14**Age (ref: less than 18)120-0.11-0.55223-0.06-0.35Older than 23-0.01-0.03Receives Youth Allowance-0.10-1.27Live with parents-0.61-8.38***Father's highest qualification (ref: no post-school qualification)Trade certificate-0.01-0.06Higher education-0.18-1.70*Unknown-0.16-1.04Father employed0.120.73ANU scale of father's occupation -0.18-1.03Father is a manager0.252.49**Mother's highest qualification (ref: no post-school qualification)Trade certificate0.040.37Higher education-0.11-1.11Unknown-0.06-0.40Mother employed0.131.23ANU scale of mother's occupation -0.05-0.30Mother is a manager0.120.70Field of study (ref: education)Science-0.34-0.68IT-0.86-1.44Engineering0.240.35Architecture1.860.87Agriculture-1.29-1.77*Health-1.00-2.54**Management and commerce0.050.15Hospitality-1.35-1.68Social studies and arts-0.87-2.69***Highest qualification (no post-school qualification)VET0.030.23Bachelor0.151.43Higher education0.513.08***Hours worked in last year of study (ref: did not work)0.1160.040.3016.1 240.060.4724.1320.110.72More than 321.017.03***Year since leaving education (ref: first)second1.3311.20***third1.6311.50***Type of job in last year of study (ref: did not work)Work in career job found prior to study0.742.80***Work in casual career job found while studying1.574.45***Work in ongoing career job found while studying1.295.29***Work in a non-career job1.596.63***InteractionsWork in career job found prior to study x Year 2-0.55-1.44Work in career job found prior to study x Year 3-1.10-2.72***Work in casual career job found while studying x2-1.20-1.97**Work in casual career job found while studying x3-1.54-2.94***Work in career job found prior to study x 2-0.76-2.00**Work in career job found prior to study x 3-1.39-3.76***Work in a non-career job x 2-0.83-2.97***Work in a non-career job x 3-1.25-3.76***Years of prior employment0.197.12***Years of prior employment x Work in non-career job-0.13-2.10**Years of prior employment x Work in non-career job x Year 2-0.06-0.78Years of prior employment x Work in non-career job x Year 30.050.40Higher education x Work in casual career job-0.68-1.67*Higher education x Work in casual career job x Year 20.861.16Higher education x Work in casual career job x Year 30.620.90Higher education x non-career job-1.50-5.85***Higher education x non-career job x Year 20.923.00***Higher education x non-career job x Year 31.413.86***Higher education x Years of prior employment0.091.46Higher education x Years of prior employment x Year 20.050.59Higher education x Years of prior employment x Year 3-0.06-0.452.1047.59***(1.3923.55***Notes: ***Significant at 1%; **Significant at 5%; *Significant at 10%.  At the time the data were collected, there was no loans scheme (like HECS available to higher education students) for VET students to defer the payment of fees.  It is important to note that the estimated higher completion rate for VET only applies to the sample under consideration: those 25 and under who are not enrolled in an apprenticeship or traineeship.  The few students who study part-time and who do not work are omitted from the study because their number is too small to warrant detailed analysis and their circumstances are likely to be quite different from the typical part-time student.  Although we control for academic ability in the form of university entry scores (or combined Year 9 numeracy and literacy scores if entry scores are missing), we cannot rule out the possibility that those who find career jobs while studying are better-performing students and who are therefore more likely to complete.  The 1995 cohort was surveyed annually from 1995 to 2006; the 1998 cohort has been surveyed annually from 1998 to2008.  We acknowledge that, by retaining these short spells of study, there is a risk of error in the working arrangements while studying. However, we consider the bias from the errors on the analysis to be less potentially troublesome than bias from omitting the short spells.  Ideally, we would examine the effect of work and study combinations across all courses undertaken. In such an approach, multiple courses would be treated as repeated observations for the same individual and estimation could be carried out using panel data models, such as random effects estimation, which allows for correlation between observations from the same individual over time. However, such a technique could not be employed in this case because there are too few individuals with repeated observations, especially for higher education students. Therefore, the results presented on the effects of combining work and study may not be representative of the effects for all VET students aged 25 and under. It is possible that the effects may be different for those who are in subsequent courses, for example, because they have a stronger attachment to subsequent courses.  Average hours worked in all periods of the course, including periods in which the student reported not being in work.  The rate of part-time study for our sample is below the average for all VET enrolments reported by the ϲʿ¼ (ϲʿ¼ 2009). ϲʿ¼ reports that, in 2002, 84% of all students aged between 15 and 24 are enrolled part-time. After consulting with ϲʿ¼, it is concluded that direct comparisons between the ϲʿ¼ part-time study proportions and the proportions estimated from LSAY are not possible. In the ϲʿ¼ data, part-time students are those whose scheduled course hours are below 75% of the normal full-time study load of 720 hours per year (540 hours per year). However, the scheduled hours of study is based upon a standard measure of the hours required to deliver given course modules, which in reality may differ significantly from institution to institution and may not represent the actual hours delivered. Part-time study in LSAY is self-reported and may better reflect the actual VET student study load. However, it is unclear how students identify themselves as part-time or full-time and whether they do it in a consistent way.  In the case where study is begun and completed between periods, enrolment and course outcomes are observed in the same survey.  Therefore, caution should be exercised when comparing the VET completion rates reported here to those reported for youth elsewhere, such as in Stanwick (2005), which does not exclude module completers.  The MNL model estimates separate coefficients for each work and study combination, which means that we need a large number of observations for each alternative. When examining education outcomes, work and study combinations are an independent variable, which enables us to examine impacts for more combinations.  The ANU3 scale is a measure of occupational status that reflects differences in occupational bargaining power, prestige and rewards (McMillan & Jones 2000).  In some years of the survey, information is available for the Queensland Overall Position (OP), but not recorded for the Transfer Index. To convert the Queensland OPs into Transfer Index values, we used conversion factors provided by the Queensland Tertiary Admissions Centre. Because there are a number of missing observations for this variable, we include a dummy variable to control for potential bias from the missing data. Observations may be missing if an individual could not recall their score, or if they did not receive an entry score.  We attempted to include interaction effects to test for this, but possibly because of high multi-colinearity among the interaction terms, the model would not converge.  Where they do not, the discrepancy is due to rounding error.  The few students who study part-time and who do not work are omitted from the study because their number is too small to warrant detailed analysis and their circumstances are likely to be quite different from the typical part-time student.  Exclusion restrictions are facto12DJST    < = T U X ^ a w    4 7 H I J K L k m n o txy쾷𪠒 h*hWh<hWaJh/\hW56B*phhW5B*phh/\hW5B*ph hy hW h(9hWjhyhWU h-NhW hyhW hshW h[chWhWjhWUmHnHsH u323456T J K o $]gdW $]gdW$]gdW $x]gdWgdW$$]^a$gdW9gdWC]gdW 'h]gdW']gdW&]gdW9gdWtY*Ha9gdW(]gdWgdW9]gdW 9]gdWxgdW ]gdW$&dPa$gdW$gdW $ gdW$]gdW)*[PGHRS !ϻ|v||ooocjhWOJQJU ht2dhW hWaJhMhWaJ hMhWh{PhWmH sH hWmH sH hzhWaJhhW6CJ aJ hW6CJ 'jhi|hW6CJ UmHnHuh/\hWB*phh/\hWB*CJph h]hW h*hWhW h-NhW& W=|8x? >W0W9W!/078789;<=ձձձ~ձmձ jhWUmHnHu jwhWUmHnHu jhWUmHnHu j}hWUmHnHu%hWCJOJQJaJmHnHsH u jhWUmHnHujhWUmHnHuhWmHnHujhWOJQJUhWOJQJ*=[\vwxz{|234678WXr jehWUmHnHu jhWUmHnHu jkhWUmHnHu jhWUmHnHu%hWCJOJQJaJmHnHsH u jqhWUmHnHujhWUmHnHuhWmHnHu+rstvwx9:;=>?{|شأؒ؁p jShWUmHnHu jhWUmHnHu jYhWUmHnHu jhWUmHnHu j_hWUmHnHu%hWCJOJQJaJmHnHsH uhWmHnHujhWUmHnHu jhWUmHnHu,   89:<=>RSmnoqrs j hWUmHnHu jG hWUmHnHu jhWUmHnHu jMhWUmHnHu jhWUmHnHujhWUmHnHuhWmHnHu%hWCJOJQJaJmHnHsH u,  + , - . [ \ v w x z { | } ~ M!N!h!i!j!l!m!n!o!p!!!!Ź؜؋z j hWUmHnHu j; hWUmHnHu j hWUmHnHujhWUhWjhWOJQJU%hWCJOJQJaJmHnHsH uhWmHnHujhWUmHnHu jA hWUmHnHu- | n!!4"y""R##B$$%%&&''+'6'*-9gdWgdWgdW"  "  " I " I gdWgdW!!!!!!!!""."/"0"2"3"4"5"6"X"Y"s"t"u"w"x"y"z"{"""""""""""1#2#L#M#N#P#Q#R#S#T####شأؒ؁ j)hWUmHnHu j hWUmHnHu j/ hWUmHnHu j hWUmHnHu%hWCJOJQJaJmHnHsH uhWmHnHujhWUmHnHu j5 hWUmHnHu2########!$"$<$=$>$@$A$B$D$E$$$$$$$$$$$$$%%%%%%%%{%|%%%%%%%%%%%&شأؒ؁ jhWUmHnHu jhWUmHnHu jhWUmHnHu j#hWUmHnHu%hWCJOJQJaJmHnHsH uhWmHnHujhWUmHnHu jhWUmHnHu2&&&&&&&&&&&&&&&&&&&&''''''''+'++00G1L:U:;;;);;شأśugjhW0J0UmH sH hh&hW6mH sH hWmH sH  hwqhW h~bhWhWjhWU jhWUmHnHu jhWUmHnHu%hWCJOJQJaJmHnHsH uhWmHnHujhWUmHnHu jhWUmHnHu(-00"4t9;);>CKWWW\9WWW9ՄW9WW9i]igW;;;<T<>>>?rBsBKKWWXXX Y Y Y0Z1Z[[[F\X\\\\\]]]]]^^m`u`eeIg`gagbg,isiiiiiuu ~LJJKV hWPJh1hWPJ h rhW h#hW hu<hW h'JhW h[UhW hhW h 1hW htghWjhW0J0UhWmH sH hW@\m`cddeegiMmjoruu]x}gdW9gdWgdW9gdWF9EƀgdWLJv, $$Ifa$gdW $IfgdWgdW9gdWgdWH9EƀgdWgdW VYqr߈'Fɉbc /0LMghZ[hw hWaJh rhW5CJOJQJh rhW5hWCJOJQJhW6CJOJQJh rhWCJOJQJhWCJOJQJaJjhW0J0U hhW hICYhW h rhW hhWhW3,-.8BHR\bssssss $$Ifa$gdW $IfgdWwkd$$IfFn9!%%q%uH!    2 24apbcd# !$IfgdWkd^$$If֞n< m9!%%H!2 24agpFdjlrtz| $IfgdWFfX !$$Ifa$gdWƕɕΕѕՕؕݕFf$$If]a$gdW$$If]a$gdW $IfgdWFf "$')-/0246Ff( $IfgdWFf7$$$If]a$gdW$$If]a$gdW68:=?ACEGJLMOQSUWY[]_aceg $IfgdWFfi-$$If]a$gdW$$If]a$gdWghnsw{ǖ,gdWFf6$($If]a$gdW$($If]a$gdW ($IfgdWFf2ǖu[sn$J9iEƀ]igdWgdWF9EƀgdWFEEƀgdWFGlq34ghͤΤ23efɥʥ@G[]!$&Ľᷯᯍ hy*hWh^whW5aJ hWaJhhW6h^whW5CJh^whW5 hWCJ h^whW hWH* h$AhWjhW0J0U hqhWhWjhW0J0UaJhw hWaJ hWaJ4[ʣˣ̣֣neYYYY $$Ifa$gdW $IfgdWwkd9$$IfF*!%R%9!    2 24ap $$Ifa$gdW $IfgdWgdW kdb:$$IfH֞ hH*!%R9!2 24apF $$Ifa$gdW !#)+134@DGJMQTX[$$If]a$gdW $IfgdWFf = !$$Ifa$gdW !$IfgdW[^`dghrvy|~FfE $IfgdWFfmA$$If]a$gdW¤ŤʤͤΤڤޤFffN $IfgdWFfJ$$If]a$gdW#%(+/23?BDHKORVX\_cefFf W $IfgdWFfR$$If]a$gdWfnqsvy|~$($If]a$gdW ($IfgdWFf_[$$If]a$gdW $IfgdWťɥʥ8 $$Ifa$gdW $IfgdWgdWFEEƀgdW,gdWFf_$($If]a$gdW Ū˪ժߪssssss $$Ifa$gdW $IfgdWwkdb$$If F*!%9!    2 24ap( !$IfgdWkdc$$If ֞ V*!%%9!2 24apF 269=@DGLO$q$If]qa$gdW$$If]a$gdW $IfgdWFff !$$Ifa$gdWOSV[^_īǫʫͫFfo $IfgdWFfPk$q$If]qa$gdW$$If]a$gdWͫЫԫ׫۫ޫ $q($If]qa$gdW$($If]a$gdW ($IfgdWFft$q$If]qa$gdW$$If]a$gdW!%&K 9;h$If^`gdW9gdW9gdWEgdW,gdWFf5y$q($If]qa$gdW$($If]a$gdW -2Zdeb})*:`ʴ"*o|۵ܵ 8;hiXYSTlm>?lmh ChW5 hWCJ h ChWhc^hWCJOJQJhc^hW5CJ h^whW hthWhWLhiROd$EƀIfgdW\kd|$$If49!"  t 9!22q44 4Taf4p ?6 $IfgdW]kd}$$If0*!&&9!2 24ap $$Ifa$gdW $IfgdWgdWHkd|$$If49!" t9!244 4Taf4$,4:GMSTUWY[]_acegi !$$Ifa$gdW !$IfgdWFf $$Ifa$gdWiklm$$If]a$gdWFfT $IfgdWFf !$$Ifa$gdWȽʽ̽ϽѽӽսؽڽݽFf֒ $IfgdWFf$$If]a$gdW$$If]a$gdW "$&(*,.0258; $IfgdWFf$$If]a$gdW$$If]a$gdW;>?PRTVXZ\^adgilmtvxz|~Ff$$If]a$gdW $IfgdWFfX$$If]a$gdWþžFf $IfgdWFfڥ$$If]a$gdW$$If]a$gdWžǾɾ˾;ϾҾվؾ۾ݾ $IfgdWFf\$$If]a$gdW$$If]a$gdW !$'*,01EGIKMOQFf޸$$If]a$gdW $IfgdWFf$$If]a$gdW 01ab03568:"$AEuyz{|C12y >?;<_` h0hWhhW5 hhW h ChW h$`hW hehW hWCJhWPQSVY\^abhkmoqsux{~ ($IfgdWFf` $IfgdWFf$$If]a$gdW$$If]a$gdWF,EƀgdWFf/$($If]a$gdW$($If]a$gdW!-7>DIQYssssssss $$Ifa$gdW $IfgdW]kd$$If0*!&&9!2 24ap $$Ifa$gdW $IfgdWgdW Y_lrxyz|~ $IfgdWFfe !$$Ifa$gdW !$IfgdWFf $$Ifa$gdW$$If]a$gdW$$If]a$gdWFf4 $IfgdW !$')+-/Ff$$If]a$gdW$$If]a$gdW $IfgdWFf/258:>?LNQSUWZ]`cfhkl}Ff8 $IfgdWFfw$$If]a$gdW$$If]a$gdW $IfgdWFf$$If]a$gdW$$If]a$gdWFf{$$If]a$gdW$$If]a$gdW $IfgdWFf  "$&(*-0357;<ACEFf $IfgdWFf<$$If]a$gdW$$If]a$gdWEGIKMPSVY[_`tvxz|~ $IfgdWFf$$If]a$gdW$$If]a$gdW$($If]a$gdW ($IfgdWFf@$$If]a$gdW$$If]a$gdW $IfgdWFf F9EƀgdW9gdWgdW,gdWFf$($If]a$gdW$($If]a$gdW12gh/0ij Erw $-/?C9klݾעhY[hWCJ hY[hW hohW hy*hW hhW h^dhW hhWh0hW5CJh0hW5 hWCJ h0hW hj@PhW h9hhWjhW0J0UhW7 123i` $IfgdWwkd$$IfF*!%Q%$9!    2 24ap $$Ifa$gdW $IfgdWgdW9gdW 3=GMWag $$Ifa$gdWghi( !$IfgdWkd$$If֞ hH*!%Q$9!2 24apFioqwy$$If]a$gdW$$If]a$gdW $IfgdWFf !$$Ifa$gdW Ff# $IfgdWFf#$$If]a$gdW$$If]a$gdW!%(,/06:>BFKOTX$($If]a$gdW$($If]a$gdW ($IfgdWFf'$$If]a$gdW$$If]a$gdWX\`eij $$Ifa$gdW $IfgdWgdW,gdWFfd,$($If]a$gdW$($If]a$gdW  $.8>Dri]]]]]]] $$Ifa$gdW $IfgdWkd1/$$If\3Rq*!$D$$$09!2 24ap( DEkd 0$$Ifִ3 Rq*!$D......09!    2 24apPEFHJLNPRS !$$Ifa$gdW !$IfgdWSTkd1$$Ifִ3 Rq*!$D$.$.$.$.$.$.$09!    2 24apPT`cfilorw$8$If]8a$gdW  7$IfgdW $IfgdWwx kd;2$$Ifִ3 Rq*!$D$.$.$.$.$.$.$09!    2 24agpPx$8$If]8a$gdW  7$IfgdW $IfgdW kde3$$Ifִ3 Rq*!$D$.$.$.$.$.$.$09!    2 24agpP$8$If]8a$gdW  7$IfgdW $IfgdW kd4$$Ifִ3 Rq*!$D$.$.$.$.$.$.$09!    2 24agpP$8$If]8a$gdW  7$IfgdW $IfgdW kd5$$Ifִ3 Rq*!$D$.$.$.$.$.$.$09!    2 24agpP$8$If]8a$gdW  7$IfgdW $IfgdW kd6$$Ifִ3 Rq*!$D$.$.$.$.$.$.$09!    2 24agpP $$8$If]8a$gdW  7$IfgdW $IfgdW$% kd7$$Ifִ3 Rq*!$D$.$.$.$.$.$.$09!    2 24agpP%-0369<?C$8$If]8a$gdW  7$IfgdW $IfgdWCDkd8$$Ifִ3 Rq*!$D$.$.$.$.$.$.$09!    2 24apPDJNRW\`ej$8($If]8a$gdW 7($IfgdW ($IfgdWjkkd:$$Ifִ3 Rq*!$D$.$.$.$.$.$.$09!    2 24apPk]T $IfgdWxkd#;$$IfF?!%%V$S6    2 24ap $$Ifa$gdW $IfgdWgdW9gdWEgdW,gdW (28 $$Ifa$gdW89;& !$IfgdWkd;$$If֞ R?!%$62 24apF;ACIKQSY[aciklz{Mkd B$$If?!%62 24agp $IfgdWFfP? !$$Ifa$gdWlyz{JKrst34]bvwiOP/0de4E_`bƶƶƶƶƶƶƶƶƶƶƶƶƶhY[hWCJKH h``:hWh<_hW5mH sH hWB*CJ^JmH phsH hWmH sH  h%b7hW h<_hWhohW5CJ hWCJ hW5CJhWhohW5< $IfgdWFfD$$If]a$gdW #',0$($If]a$gdW ($IfgdWFfN$$If]a$gdW $IfgdWFfI059=AFJKst~$$If]a$gdWMkdV$$If?!%62 24agp $IfgdWFfT$($If]a$gdWFf^ $IfgdWFfY$$If]a$gdW #'+$($If]a$gdW ($IfgdWFfd$$If]a$gdW+/34Xwd ijs| $$Ifa$gdW $IfgdWgdW D^`gdW9gdWgdW,gdWFf4i$($If]a$gdW~~~~~~~~~ $$Ifa$gdW $IfgdWlkdl$$If\O9!9!a ! !$IfgdWkdul$$If O a&t9!9!((((a !$$Ifa$gdW ! $IfgdWkdVm$$If O a&t9!9!((((a $IfgdW 7! $IfgdWkdEn$$If O a&t9!9!((((a79<?ADGILO$$If]a$gdW OPk! $IfgdWkd4o$$If O a&t9!9!((((akmpsux{}$$If]a$gdW ! $IfgdWkdp$$If O a&t9!9!((((a$$If]a$gdW ! $IfgdWkdp$$If O a&t9!9!((((a$$If]a$gdW ! $IfgdWkdq$$If O a&t9!9!((((a$$If]a$gdW ! $IfgdWkdr$$If O a&t9!9!((((a "$'),/$$If]a$gdW /0K! $IfgdWkds$$If O a&t9!9!((((aKNQTVY\^ad$$If]a$gdW der! $IfgdWkdzt$$If O a&t9!9!((((arux{~$$If]a$gdW ! ($IfgdWkd[u$$If O a&t9!9!((((a$($If]a$gdW ;!,i]igdWkd >>>>>>>!>">'>(>,>->3>4>8>9>?>@>A>S>T>X>Y>^>_>d>e>j>k>p>q>u>v>|>}>>>>>>>>>>hdhWH* hWCJ h__hWhW[========================<$If^gdWFf <$IfgdW <$IfgdW==> >>>>">(>->4>9>@>A>O>T>Y>_>e>k>q>v>}>>>Ffw <$IfgdW<$If^gdWFf|>>>>>>>>>>>>>> ????%?+?3?9???E?M?Ffm <$IfgdW <$IfgdWFfr>>>>>>>>>>>>>>>>>>>>??????$?%?*?+?2?3?8?9?>???D?E?L?M?N?e?f?m?n?r?s?y?z?????????????????????????@@@@ @ @@@@@@ @!@0@1@6@7@<@=@D@E@I@ hWCJ h__hWhW^M?N?`?f?n?s?z???????????????????Ffc <$IfgdW <$IfgdWFfh???????@@ @@@ @!@,@1@7@=@E@J@Q@V@[@`@e@FfY <$IfgdW<$If^gdWFf^I@J@P@Q@U@V@Z@[@_@`@d@e@f@}@~@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@AA A AAAAAAAA A%A&A+A,A1A2A7A8A9AHAIAMANASATAYAZA^A_AcAdAhAiAmAnArAsAwA hWCJhW h__hW^e@f@x@~@@@@@@@@@@@@@@@@@@@@@@FfO <$IfgdW<$If^gdWFfT@@@A AAAA A&A,A2A8A9ADAIANATAZA_AdAiAnAsAxAFfE <$IfgdW<$If^gdWFfJwAxAyAAAAAAAAAAAAAAAAAAAAAABBBBBBB$B%B*B+B0B1B8B9B=B>ٵѵٵBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBDBIBNBOBhBmBtBzBBBBBB <$IfgdWFf1<$If^gdWFf6 <$IfgdWBBBBBBBBBBBBBBCCC?C@CACBCCCDCECFf' <$IfgdWFf, <$IfgdWECFCGCSCYCaCfCkCpCuCzCCCCCCCCCCCCCCCCFf<$If^gdWFf" <$IfgdWYC`CaCeCfCjCkCoCpCtCuCyCzC~CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCDDDD D DDDD+D,D1D2D6D7D;D?䷡ٷʷܷڷ췡跡ܷ<$If^gdWFf <$IfgdW <$IfgdWFfEEEEEEEEEEEEEEEEEEEEEEEF FFFf <$If^gdWFf <$IfgdWFFF)F*F1F2F6F7F;F$$If֞ J*!$-$$$$$$9!2 24apFeeeeeee  $IfgdWeef*! $IfgdWkd?$$If֞ J*!$-$$$$$$9!2 24apFfffffff  $IfgdWeffffff=f>ftfufffff4g5gsgtggggg#h$hbhchhhhh i icidiiiiiijjjj5j[j\jjjkkPkQktkzk{kkkkkkkl l^ljllllll@mBmmmmmm n nQnRnSn hChW hWH* h/|hW hWaJ hEhWh.ghW5 hWCJhWQfff*! $IfgdWkd@$$If֞ J*!$-$$$$$$9!2 24apFf f f f f ff  $IfgdWfff*! $IfgdWkdB$$If֞ J*!$-$$$$$$9!2 24apFfffffff  $IfgdWff7f*! $IfgdWkd*C$$If֞ J*!$-$$$$$$9!2 24apF7f8f9f:f;ffFf* $If^gdWkdFD$$If֞ J*!$-$$$$$$9!2 24apFFfOfXf`fgfnftf  $IfgdWtfuff* $If^gdWkdbE$$If֞ J*!$-$$$$$$9!2 24apFfffffff  $IfgdWfff* $If^gdWkd~F$$If֞ J*!$-$$$$$$9!2 24apFfffffff  $IfgdWffg* $If^gdWkdG$$If֞ J*!$-$$$$$$9!2 24apFg gg g'g.g4g  $IfgdW4g5gMg* $If^gdWkdH$$If֞ J*!$-$$$$$$9!2 24apFMgSgYg`gfglgsg  $IfgdWsgtgg* $If^gdWkdI$$If֞ J*!$-$$$$$$9!2 24apFggggggg  $IfgdWggg* $If^gdWkdJ$$If֞ J*!$-$$$$$$9!2 24apFggggggg  $IfgdWggg* $If^gdWkd L$$If֞ J*!$-$$$$$$9!2 24apFghhhhh#h  $IfgdW#h$h6h* $If^gdWkd&M$$If֞ J*!$-$$$$$$9!2 24apF6h>hFhOhUh[hbh  $IfgdWbhchhh* $If^gdWkdBN$$If֞ J*!$-$$$$$$9!2 24apFhhohvh~hhhh  $IfgdWhhh* $If^gdWkd^O$$If֞ J*!$-$$$$$$9!2 24apFhhhhhhh  $IfgdWhhh*! $IfgdWkdzP$$If֞ J*!$-$$$$$$9!2 24apFhhhhhi i  $IfgdW i i7i*! $IfgdWkdQ$$If֞ J*!$-$$$$$$9!2 24apF7i?iGiPiVi\ici  $IfgdWcidiwi*! $IfgdWkdR$$If֞ J*!$-$$$$$$9!2 24apFwiiiiiii  $IfgdWiiii*!! $IfgdWkdS$$If֞ J*!$-$$$$$$9!2 24apFiiiiiij9kdT$$IfֈJ*!$ $$$$$9!2 24ap< $IfgdWjjjjjjj $IfgdWjj5j* $If^gdWkdU$$If֞ J*!$-$$$$$$9!2 24apF5j;jAjHjNjTj[j  $IfgdW[j\j|j* $If^gdWkdW$$If֞ J*!$-$$$$$$9!2 24apF|jjjjjjj  $IfgdWjjj*! $IfgdWkdX$$If֞ J*!$-$$$$$$9!2 24apFjjjjjjk  $IfgdWkkk*! $IfgdWkd:Y$$If֞ J*!$-$$$$$$9!2 24apFk)k3kFfZ $$Ifa$gdWFfɍ $IfgdW  q$IfgdW>CHMRTVXY Ff| $IfgdWFf $$Ifa$gdW  q$IfgdW"*2:;} $ ($IfgdWFf $IfgdWFf   q$IfgdW$,289  0XS9gdWJ9iEƀ]igdWFEEƀgdW,gdWFf=  q$IfgdW 0HY)*8?EBgdWgdW9gdW9gdWgdWJ9Eƀ]gdWB~ 4 ! $$Ifa$gdW $IfgdWgdWgdWgdW B7M! +QRSmno@!*4 '26AEP^imx|ҽҽҽҽҽҽҽҽҽҽҽҽҽ h*hW h\hW h,chWhWhWOJQJh*bhW6 h*bhWh*bhW6aJh*bhWaJGSJ=0  S$IfgdW  $IfgdW $IfgdWkdv$$Ifqrw9!$$$$9!22 q44 qaqp2F=0  $IfgdW $IfgdWkd[$$Ifqrw9!$$$$$9!22 q44 qaqp2  $IfgdW   90 $IfgdWkd2$$Ifqrw9!$$$$$9!22 q44 qaqp2  $IfgdW  S$IfgdW   $IfgdW  S$IfgdW  $IfgdW%',SJ=0  S$IfgdW  $IfgdW $IfgdWkd $$Ifqrw9!$$$$$9!22 q44 qaqp2,/2346F=0  $IfgdW $IfgdWkd$$Ifqrw9!$$$$$9!22 q44 qaqp2  $IfgdW6;>ABC90 $IfgdWkd$$Ifqrw9!$$$$$9!22 q44 qaqp2  $IfgdW  S$IfgdWCEJMP  $IfgdW  S$IfgdW  $IfgdWPQ\^cSJ=0  S$IfgdW  $IfgdW $IfgdWkd$$Ifqrw9!$$$$$9!22 q44 qaqp2cfijkmF=0  $IfgdW $IfgdWkde$$Ifqrw9!$$$$$9!22 q44 qaqp2  $IfgdWmruxyz90 $IfgdWkd<$$Ifqrw9!$$$$$9!22 q44 qaqp2  $IfgdW  S$IfgdWz|  $IfgdW  S$IfgdW  $IfgdWSJ=0  S$IfgdW  $IfgdW $IfgdWkd$$Ifqrw9!$$$$$9!22 q44 qaqp2F=0  $IfgdW $IfgdWkd$$Ifqrw9!$$$$$9!22 q44 qaqp2  $IfgdW90 $IfgdWkd$$Ifqrw9!$$$$$9!22 q44 qaqp2  $IfgdW  S$IfgdW  $IfgdW  S$IfgdW  $IfgdWSJ=0  S$IfgdW  $IfgdW $IfgdWkd$$Ifqrw9!$$$$$9!22 q44 qaqp2F=0  $IfgdW $IfgdWkdo$$Ifqrw9!$$$$$9!22 q44 qaqp2  $IfgdW   UW^`   78OPbc89WXtu·· hLShWhihWnH tH hWnH tH hhWH* hhW hWCJ hDhW hhWhW h*hWhWCJaJE9. ($IfgdWkdF$$Ifqrw9!$$$$$9!22 q44 qaqp2  $IfgdW  S$IfgdW ($IfgdW S($IfgdW ($IfgdW zSNE<0 $$Ifa$gdW $IfgdW.].gdW,gdWkd$$Ifqrw9!$$$$$9!22 q44 qaqp2$$If]a$gdW $IfgdWekd$$Ifl0|9!$ $}9!22 l44 lalp$$If]a$gdW $IfgdWekd$$Ifl0|9!$ $}9!22 l44 lalp$$If]a$gdW $IfgdWekdV$$Ifl0|9!$ $}9!22 l44 lalp$$If]a$gdW $IfgdWekd$$Ifl0|9!$ $}9!22 l44 lalp|$($If]a$gdW ($IfgdWekd$$Ifl0|9!$ $}9!22 l44 lalpUVzz $$Ifa$gdW $IfgdWgdW,gdWekd*$$Ifl0|9!$ $}9!22 l44 lalpwj]  $IfgdW  7$IfgdW $IfgdWkdԶ$$Ifq4Fi9!$ $k$k9!    22 q44 qaqpxk^  $IfgdW  7$IfgdW $IfgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqp  xk^  $IfgdW  7$IfgdW $IfgdW}kda$$IfqFi9!$ $k$k9!    22 q44 qaqp  tgZ  $IfgdW  7$IfgdW T$If^TgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqp (.7tgZ  $IfgdW  7$IfgdW T$If^TgdW}kdϹ$$IfqFi9!$ $k$k9!    22 q44 qaqp78@FOtgZ  $IfgdW  7$IfgdW T$If^TgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqpOPX]btgZ  $IfgdW  7$IfgdW T$If^TgdW}kd=$$IfqFi9!$ $k$k9!    22 q44 qaqpbcpvxk^  $IfgdW  7$IfgdW $IfgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqpxk^  $IfgdW  7$IfgdW $IfgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqpxk^  $IfgdW  7$IfgdW $IfgdW}kdb$$IfqFi9!$ $k$k9!    22 q44 qaqpxk^  $IfgdW  7$IfgdW $IfgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqpxk^  $IfgdW  7$IfgdW $IfgdW}kdо$$IfqFi9!$ $k$k9!    22 q44 qaqptgZ  $IfgdW  7$IfgdW T$If^TgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqp)/8tgZ  $IfgdW  7$IfgdW T$If^TgdW}kd>$$IfqFi9!$ $k$k9!    22 q44 qaqp89KQWtgZ  $IfgdW  7$IfgdW T$If^TgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqpWXhnttgZ  $IfgdW  7$IfgdW T$If^TgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqptu~tgZ  $IfgdW  7$IfgdW T$If^TgdW}kdc$$IfqFi9!$ $k$k9!    22 q44 qaqpteV ($IfgdW 7($IfgdW T$If^TgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqpxxk  $IfgdW $IfgdW}kd$$IfqFi9!$ $k$k9!    22 q44 qaqp vi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp>?noVWuvZ[wx67KL[\rsVWefyz  &'STij~*+h hWCJhWa17>vi\  $IfgdW  7$IfgdW T$If^TgdW{kd;$$IfqFi9!$ $k$k9!    22 q44 qaqp>?bgnzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpnozm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kdT$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kdm$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kd $$IfqFi9!$ $k$k9!    22 q44 qaqpTUVzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpVWiouvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpuvvi\  $IfgdW  7$IfgdW T$If^TgdW{kd9$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kdR$$IfqFi9!$ $k$k9!    22 q44 qaqp zm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpXYZzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpZ[mrwvi\  $IfgdW  7$IfgdW T$If^TgdW{kdk$$IfqFi9!$ $k$k9!    22 q44 qaqpwxvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kd7$$IfqFi9!$ $k$k9!    22 q44 qaqp zm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp456zm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp67?EKvi\  $IfgdW  7$IfgdW T$If^TgdW{kdP$$IfqFi9!$ $k$k9!    22 q44 qaqpKLOU[vi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp[\hmrvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqprsvi\  $IfgdW  7$IfgdW T$If^TgdW{kdi$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd5$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpTUVzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpVW[`evi\  $IfgdW  7$IfgdW T$If^TgdW{kdN$$IfqFi9!$ $k$k9!    22 q44 qaqpefotyvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpyzvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kdg$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp vi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp  &vi\  $IfgdW  7$IfgdW T$If^TgdW{kd3$$IfqFi9!$ $k$k9!    22 q44 qaqp&'QRSzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpST[`ivi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpijpu~vi\  $IfgdW  7$IfgdW T$If^TgdW{kdL$$IfqFi9!$ $k$k9!    22 q44 qaqp~zm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp"*vi\  $IfgdW  7$IfgdW T$If^TgdW{kde$$IfqFi9!$ $k$k9!    22 q44 qaqp*+[`hvi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqphivi\  $IfgdW  7$IfgdW T$If^TgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqphi_`EFrs%&lm)*[\FG "$BD ɼh=QhW0J: hW0J: hPhWjhPhW0J0CJU hhW jshWhWOJQJhhW5hW hWCJFzm`  $IfgdW  7$IfgdW $IfgdW{kd~$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kd1$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpQW_zm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp_`zm`  $IfgdW  7$IfgdW $IfgdW{kdJ$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp zm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp6<Ezm`  $IfgdW  7$IfgdW $IfgdW{kdc$$IfqFi9!$ $k$k9!    22 q44 qaqpEFcirzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqprszm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kd|$$IfqFi9!$ $k$k9!    22 q44 qaqp%zm`  $IfgdW  7$IfgdW $IfgdW{kd/$$IfqFi9!$ $k$k9!    22 q44 qaqp%&bglzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqplmzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kdH$$IfqFi9!$ $k$k9!    22 q44 qaqp$)zm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp)*LR[zm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp[\zm`  $IfgdW  7$IfgdW $IfgdW{kda$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpzm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp<AFzm`  $IfgdW  7$IfgdW $IfgdW{kdz$$IfqFi9!$ $k$k9!    22 q44 qaqpFG}zm`  $IfgdW  7$IfgdW $IfgdW{kd-$$IfqFi9!$ $k$k9!    22 q44 qaqp zm`  $IfgdW  7$IfgdW $IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqp "&0BxiZ ($IfgdW 7($IfgdW ($IfgdW{kd$$IfqFi9!$ $k$k9!    22 q44 qaqpBD  {8~yypyyyyyyy-]gdW-gdW,gdW{kdF$$IfqFi9!$ $k$k9!    22 q44 qaqp   {|}0167ag~ai89?@AH}v}nv}heUhWH* heUhWjheUhW0J0CJUjh0hW0J0CJUhumnhW0J:h^whW0J:h^whW0J0CJH* h^whWjh^whW0J0CJUjhqhW0J0CJUjh rhW0J0CJUhW hW0J:h=QhW0J:+8?+^_`bcefhiklm $dNgdWgdW-gdW+,- )]_`acdfgijlmntuwx{u{j{c{ hPhWh_}0JmHnHu hW0JjhW0JUjhUhh=QhW0J: hW0J:jhIhW0J0CJUjhSXrhW0J0CJUUjhhW0J0CJUjh]UhW0J0CJUjhChW0J0CJUjh!yhW0J0CJUhW&rs that affect hours of work, but not the likelihood of completion. Valid exclusion restrictions are important for the identification of the model. Both of these factors were found to be significant in the choice of work hours, but were not significant in the completion model.  Predicted probability of employment in the first year out given that the student worked in a casual career job in their last year of study (86%) less the predicted probability of employment in the first year out given that the student did not work in the last year of study.  Although we control for academic ability in the form of university entry scores (or combined Year 9 numeracy and literacy scores if entry scores are missing), we cannot rule out the possibility that those who find career jobs while studying are better-performing students who are more likely to complete.      PAGE 42 Outcomes from combining work and tertiary study ϲʿ¼  PAGE 43 ŏhhW hPhWh_}0JmHnHujhW0JU hW0JÏďŏ,gdW;0 &P /R . A!"#$7%0 50 &P :pW{. A!"#$7% `!YeŌ,uBTrll,9xX pU]k(GBETZD! H@1P$EJi@Bak0VtpZ*uZ,FNNcIZd> ZdN-2g'u9sIs3}z"'2gG#zbAA=Ey=ѠE9=Elx"sVȜ׌ZdVh-2g՚1AYfP"sVȜh}h0AYwؠE欃fp"sgs"sgw"v5/2g2F-EElVQ"sS*#ەζiȜjU-2g+JEZAYVq"svVu"svVy"3֪Zd -2gnAYGuE̩-2gU}3}luS]Ȝ%եY:@2dG-0B/oNSw*Yk݊p/6m6\niKeo1Qc`@Dc=(nO Qb{4 3l/l{245,#Xlce'V k1Xk'C{?6Y9l`qV+~=Fnz־u.R_kXns9s=RtP(Pos=đzKUz6+JB1Me{¯Xmm2M8IL W 9DL`dWw1˄ QP- PlrS'ɏ9ZB?T9gxO/<^<ƞ|؃^bVTl;pVu9k95:bZX TpyqܢxPmxTN(l~-V*JX-m-{i%6q!lv ϲ{6(L/L|)^b_r!^\w,ÛJѐL ;')ߊ7 O3qIY`sIwWsf?QMTn8Vн ŰPNG  IHDRt pHYs,J,JwztMIDATx^Ui PB @[Nί bcv7a (H(;{O̳ԙΙse0 C0 C0 C0 C0 C0 C0 CXhݺu%Q C0 C0 CXjԴt)hh͚5ϮSrϟ4ךh!`!`K-aYjs饗.x]vٞK-pC0 C0 C,[73*{e5j4_fJyJ+5^mՆΚ5ߔy,!`!`!`APKW X潇RV.IICE8 i%oy C0 C0 C(OA-Ot˨W_}%[8wz|iAr`-Җi C0 C0 C(  jYP.a:6mڴeiڵkM+Qť)Ky)Gj'Y;IDAT@[Җk C0 C0 C(-&A-5\sڤIU9+*U7VUd50ukԨўrh A ~41?vĈ=V!`!`!`e1E xNJoVgʔ) `Et0YMtϗ\Μ^@Kd!`!`@%"`*ECpߐp5͢J}vOqUA,xfhlXL0 C0 C0 B"G"]4 :4X=HM_EVQL(:ߝ|H!q&-!`!`!`1EB/0_ƲՂQ&v =a<'YE0#/o7Di&u0 C0 C0 JBԒ/<2iɪH!H؆s&QTG jf͚4U+,!`!`!` ` j A޽0MV-/2ǙrejS T~yZu)VuYYEs1*\i4!`!`@!P^K;w:ulNKۤhm (1)V$񮃻W :N5Z5Έ5 C0 C0 C$C@RD4 E_f-~ҤWo\Is]X 'PW^_Zz:uĉO~QyZ Crs^IDATEM0 C0 C0*cPKk2_)"5bR1^t?+Tu׮]k=zߪ.W\IxiӦbM²(qhg0`4u$4ws+ΜA\SYrp_!_Y/|;,_%!`!`@[] JnFޛ9UI0pa*?9<V0rIr$]}=Sxnn͛7ogˆcivĎHɞ@ `rs_z@i|}VI|!`!`!d#PsdCQ|`2@JX p_,>/3uF}Ir+))o+c:e?sYϭ$}ɤ_~QɄGD &-ӧO풪5V!`!`!`T-A-E]OsJQ@$JfQMDa!`!`!Po)GآiyCzӍ=̩H:o4!`!`@ ` jpWaNC%_#6K.C0 C0 C0o)aCILٳH ~˙Y?uԙfEy=E)*#,F!`!`!`T*Ơ~1nXťZei}7tm/V 2b C0 C0 C0AsI`L.e1 21l0 C0 CX0=jHּ>Da }jYfiyWMcPC0 C0 CA-!=%Eeo1K?K#$4tn_[A!`!`@0!=mMi"Ul&]uNYiy?[unn!`!`叀1%f<5 !nIDAT.VګV5\Sj7Y2Z5!`!`!PHdU9+MA=;tr%Hf [NR-ڊI_ yjCMB]>%[!`!`!E"g-/$KA3,fv)SlTR`ΪιIٳWs[jb$VXr۟*P`\L0ʡ"EcRX+0 C0 C!P j&M:W&խ7=4W,݋-&M6V_}C1QFtyoHvܹ#IeuwuYgȫK_`_0a_0 ѣGCOJ++BZ~&l51_mmQϟ:6lʸqf3}J2v!-0 C0 C0 #PiR`Q0[0;R`NZj4҈#1tQ&Mf1G/ ЙP/Bi:?sz 'R&]'mƑ6Җwύ`5BXX!`!`!`C$!*#}z6 9z+IPP08kWIDATWpR˂vEiI@'Hj$-ޡ 1,s)U${7p}v,Ϯ]y~,Qa!`!`r ]8˯+ܪ3HY1a[%U@⹋p|1UI@^<8F~/A08F6+s#{IKs47siVZIW:01] Sz.sZzh4 C0 C.T0_cazP,ЙD,B/"X^zrv,K$A*!OEaxISIs/*g ]m鿦2{S0S|fG?}O?A@$-)w;Թ1c:ǁ$mCHyhGʾIVb4!`!`Ur6j?=4X2*=U \ =eɜʰP1`vAx^PFuuMt'tW8e AK s %c=5^) `q?^x~$`F C0 C0 C @3cǎPiOC4iC㔆`FBղ +Wv)ܒ6!`/NIDAT!`KΠ=Z QӠA+`rfWHP=-XwEbՑ_E х:wLIYU}?}řh"&Th&KJ4oRoTZݐr\YȧgA *ˡIq$o͂9z+$H鏄_h<7ʸ Zd+)=xGh\3ɳy6PFpCԇɱy. .jE{ ZV*L4U#6p %L$#7GO"R,W>ہBI%j$܏;x|WEyUe_"Efz\5w{I=Jږ` pU-ؐylޥc}ߤc \CޝuO2&TKʈhW`z̧ӽ\(' \Icc29et3?Y2Ɋ1 C0ʝA_X)>m+򡖤Og6Y }o\]=rbNEU觀t$3r7BEZhqxfu91't o4azd Leoy%cjIDAToM~c\zT5>ڦL1Ũ=W0ړ"Ϊųݴʐɹ9|< :62_)ݍ9!}њ a޽{7 ]'$F>1(Cꘉ'ij[!+1 @2 \:St#Gdz) .v ?|*OX..D*hp, A .Az>4 ͛7+ \ ~/1,ys I&)s#qO~Zt%H:g:?(W;:u< TRD V j_RuC1NvvU 6oJ5&P]V<d.jMƄ<Y\+{m$f1_jm.5?߯_9ӂo.[biCޛ=!m0*rgPNn;2i܆A(.9?&ޒ^ȳܲ<0†l1i-- ҞG#MKc,̞=|=,[Z#P!,v_ Fm#DsZX<ХX<Z2ffr? l' Ɗ`':izhIs( #a6G7k20US52`?%E&A mM JY2MoIGDOLR3 nP܌1[疓紷8n*3IDAT%I#iDڙ} asKZ"QR땫,_r0_SLi!->(R!i GL4MY(f Fl9RS%F!6j g$8[ +uƕݙ@$ ub/#ٙOAR^iu2\gne*|=}SƉx Pk`uF'T]0o|pxְZ,h>>d0ӹ_ܙV1MY$_z2uyR{wvtsG4JT&`>? ֍x7bguKydw!Ю\^YZp IuxoT[4 l$e2eP/f F)| y,b¤&%ݟ4iVuPEK z^%,%q1+^J @z325Ӣ$HCea}rJk~X l~04d`9X+$wSqb.3Yx+n1N`P]{o5k֎.ݜ ZTyvCDi,,mQIL+ϥysd6*`xIOZb}`hIDAT2CM5A@3 >24$0TeGIhMշ,2d޿mY# Cl\a>!iKf,bgd`8mڴ!$?ө_S.Q7۲Pܿ1ZgZ|%IKO&m|m #Dѳ q~+{A]$ou֕E0aԁ tJaaw7C{a[Nj-[V'UI#I }%p4o\>s YUoIJ&s8b̥%sGL9˛Җ,1۬Y6X-$*/R~>oZcljLu騄h@oxRRI^%lժU"Eu 'Yᳲe-4c>6O<;%+X@ZjmAb0VxpnzʴU].)a>eѴh%mU뿖3 )͡ҤϗFeeDXP~iFژ#A|_ZPoT[IqFTcw$ :S &aI$U0MK/ziN^6}isn -~9S*&@ijM-:g\ZO{O%zܹs [-ʀjq̘1'5 Rrk2̑Qơ'$y, u씦YM pm?": ?i/'(6D}6K>}ܖVA #eȂޡFzJթҔK[cxh@\o3r ʢlH }W`^d,<.3 G\ gp: a_&ܧ`X H&k4t =!ӪF:u~KĤבz[t O&բbX&R>CR1Z~+Xyw''F 8RNЏ|0rT>J0w|D/x%a+1^"M]C'kR IcTV! S,g;F❔1)rfPahw΢:|p0-4y##Ŕj\k^"8si1#UnSFs˃Ax6^-wJho o-p}mcu$H01ohs^@Wڮwvjm_QVp!1/ʘ&sd}H_sOB|H?B#Mysc=c>]\dk0Ms u2imݺuBg;p9Q x\l\bdcILRyǫI=]Koc[|s9GkTc2#(?1y/ IM=p6'N a$!һ/n>-0n]c\sbKyGI5v#4CH:҅uIDATo9^;摯^lyEگYWu UlUbND(ڿ,e]]ZCicF1.7iLX$<;*C&6Ωin&x Fhɠpl$ay(u.~̋Ȓ@Au*ڱ; u]&"_՗D40_=TζMѲeWp-"S}<Sш,Ii"/tG:|($f㏁{bN}0K-vGY-b*0U!cD,X2^)dhXv@~i`NvXL}dR?nIv #0FTgx3Gz.$0RNqb)])JkAx/WM]pGCs9%`51OY o`u9v=y0E>x,TZz̊N®}Wr? ugUŸ%D]%]Y ۈs䷵.sS||q0#5o˚9=SulЈ0` hGgb\fu}}28?{χ$}eU=e]$!dN^y9%GW}71睙މ6dI7#,];F =Q_QzHc\/,u2cGSrm.IGXy.CIPh\p޼y&%%zo J&TmA"tSi'tm?VIiͅn.nUVzu~[i{*͇mlOkgb|̱|c Cr(sU;L |,Td`+yĉcBzdTUI5J /iCh D)سT`f2{yũN ]T%UzN.%ძـ+IDAT|8,F\~: \bDI܇F:ʬ)ŃΌA-2Y"HA}CcЕ߁j.']Gҍsy_,(SXN5hzTcv-uݚ 9%MER[,#^Bixic>,‹6< *'kcm0tᝅ0B9ehE!l9OV2 DT}XPHZ><sL1Z$0a>]g"`؃p2RH3|RY]wluFR„ H%sB"'$%..t}xY<K[IuC09+%Sf ]R/I#|oE˲tF˻^ʈg ^κdT!Ӣc2Ja$ Mhʯ2apDsRGzh_YIt=x<'f@e^8)bư˛JM?RR*ƽ:*YRb:')?tAczzj"+]ڛ<3*j hB7-#F-2?Fl;pP 51(A CFNR*vIDAT`kµMR,<,~̦@_6*?7>n%`u/gĨ'd N&똴k'URk#=(KzOmH}M{%1saܵ N:!TOt%(Vs6ub74d.d8I%Xqi|2iQP.f͢h 5%kq&IT 2^MϤNFJ| %0<+ 4W!_&+9sx^ *zUc:o;w.I/p^ 0s]IWAb.dѻ@'Ψ<\T} K)X}>ڒǡ!9Um'F%u*f,|(99װ"6t&Tgs.}'0 ?'?Ͽ*9BS=$6:M*}.gtIj>o|юyP &Zw惚<aWAŔGpRߖv]8S.a E >4E v}+ny*IbS}WpȻA&,)y'iPeAf9sdT*aA wYR_ϷZ:3t^!}nNgG3uɪ;μW};;m3.o/X}_h".I~yrhȓ^ůCFc^HbU5IzR茳3/А5ߥ/Kg2ePOp鄭!QzE8h$5XW3 /z&*v !I\T&̪L;Y21h{ݤ| 'h}XD\4]:-~2^ I:6}<^˗vʰ}=$ e\KŘ3+4>;б#҅󴳴^c<Xe 1^ hoZ3fc2$k5E95k,E UlXs2RXiO̓Q#\X-׹h:WB)M$F7MC˞B$QɒV嫀~+)'I[=&u;Q"\eb|Ke=҆HR>2<O4$w8iiM(ĉy\9SB{vIX4m0eʠk!ɓguٹH@ ,!8|/˾>'b~a1U>]CXJ[H{%g0/{*8} 0K-_yˏ'4U6TkZ\=.beB!3!E?|wH|f^мIDATR:]ӗbapԗ=x8DT4}4BI+UEH#}G1|͇0LiT˰QάKr܁6hRc,muM)&. ڏ{ :DkM-3TUV'(qI]ڝnVI|Jcj }k3YƼ8Ahp؟kFq@ {DC1 gD樨EH_˰kȚI$gi'7P%e%O~ԾİH-!`T$Ơ#+:b4 bs}@8 G2v{q?#?ѴNp,`Oc($gX'eӿ/ F.6H!"Z Epe_,e0iLa>3ߤ_% B2V%xpL6@s6O6!Hū|c.kZܪ3F:sIWOQwfA죃>m#-KY~+f64 1Y&JYZF]Sp3y"sԗ+=Inc}5V< hvyd)c!SO ɛLnR]bhr&EilZ!311\kyIkbWv 24MJlHdTkI^22`T{l>E<ʙYSoEϴg6\VU ˼V!#Pa/%\LK,Z2d06#RE]xهJ"#tkiiur>cڴǸÝ˪5ڰv{ѹ4Տ̃N>2(C$5kP#mY #rV0}iK(}@HsvI+ 3QJs>3Ҥ2B|d)*Gn5*O9"%zj$p:2V i 6IDATȇMJ6G ˺OrEVUȈR$ih0W>d;;G1Zt'X݇3ԝ4viE ^^bDXhq9S1㓏Ӡa;uGBnh;L&թ/>r#142v2;3X=Pԩ9DREW . {bY%YLJH{9 'd-a'kIj)VW%AL QLEї~2)#!Qx_2oί341RsRB yM DRGsw8kAycr= _슈]ySa6f1ZdY @wXI-T|M3/8I2c˸`L%l|5kw״gI*%ڤ>\%=U4a)ㅰt%MiI;7H/Ҁ@1I2M猦Կs Ȁ {)|ȇ֑İf ďc\f]|^ l݌E$XIIa-L tA+vd&EZP~6䲇E~zbfWۈ98q{ԳI;E-BHÝ*u,>L 1:t+)7jP7,C&Z ^è~@x[u f tGz; |Fd"ԙ?!?y/z$cL0eb6YJe ~RԆ/e1fU3v :NCCݺuO YESVKډ*^2,sY":(I&C$O>4?ij?۱{2)aQܟ1mz4Dwe>] L8B+R^L9p%97uQkr_?X,'qjI8RV!$,TVbaeAS_M(KKz KerWf*Qm;5>'!Zc``,mED+S*޷:&^1UR^2KiE2 W% %4AXg I眫zBeA[z)Zmʔ['8vۏ|D+h0`Ф3mH}.mڵRM%K}me ʥ,Β!PbOFdc cTzwf1cBYƢtҟ-`x~LQ>QbD^*Ř|igzaN%_̧ܡ?}kL"4ka޽x M"3 ~wԶdh}(E8è(Vȫ*a*8wY0+Ex2U0;KƆ }4{QH $,-RV@XB _ӶS|j񥥧u^,pIPw}i.,zPǛ>ҡ&Bz;xv:r01IDAT,2f"/l{:Ha8aYR1ˡKbZUM+eڤI! lt IxU|ߕx C̩비j*s1ߨD>&Ϩ`&/޾_֓4U<5iH4!PER 󡶽K4!IRg+ÏRoc׆!9:Ҡ3鿏vz%-xQζ,3gF-ڑz".6Sؿr/&/SZXqu4F(QrL9)oj g:bBhAvc˖-k#i-?ߖƹ%j}f ߾?XUwXL7/BRTU`xzqɈqè]8(&g呻3Q)+׉t+> q8{%ԣ Eτ"IUXF&rF$#͏bZj%BF/xƚqΩeerq4eIIؐE$Kdt-kCϏa3-Ǿ԰(QAE?[h0;&$0;?@9;+iσŷZr~JB=93-;=յUZ۹#ڴ波ˆ0<r([C>͟*MpRKqi[3 CT2ЁsoCapH܉zb&e$/Kxy{UGar(&RK"&zԱ|T*[l+ 璉HKJ\wDJ-V*Ծ.r7ǂwt'1c W؇t>aV&h"|>|{Q xeF~*p=> 8+Z$d%z[TȂAOTZ:Sey@.v`w o)5|9MGunU Ͻk(c/0r Sm:I,OQIV, 5gMULERZLZIDAT}y;Lt/°Ջi‚)ȗxV#=fΣG㠡o3$1\y9csyi4EJu;9a3uHM̭]1^IJc{s@-g`g`3v@G&iGUIZQt JdթӧF'fݣgtYFsC"x6gG⿬f7¬ɰR\6 }ĭBj[rG# i.cUIU1VZQ,6gvڽJ9?^K,j}K'u^1dȐIv̯η<Bsʳb`| Eյ@B?S:Eċe14u /3c?q z/ôJ뻳wCm 17yqˆ TBcP LOk5ꔔ8 |D]йϥ^:)6!>X3=*% !;&I>:"H\"zC9ЌT>kI2GiY2>!*K=xnVȸ$w/2,K=drEtN00.v@FF%Ic[NCKe ȷDʦ@ٴtY2LzKG10-t ;vW2^zxU΄I͑[/G§ (f5039j+}R[klUdيYpaǹۓhWRLN;Zϥ彠y8ӧ>] x@1a[橌GG Iڂ[;ۮI"i%IDAT4^K}^vcIݴ4@uF3/HאfZkd’kܷP}x&y^y^9WڄŒw9rk>H߳M$ Y{qΜ9H +Lֻ*0h$x>bZ QZD#_#/OU+_'k##D-)〨4e1 hZ_;*L'ߩ{:%"nVZ$b"~WEx|4Y ~ jkxHL̳Zd-hrAЬŢm;hY)z{u$O*[>I'+c%M7([Rzg/蹅>5I׃|Oi4y~pkcHwTNuH;K7Zl I*9[NދsUY$#Eh>3_A,۴Pc:YSK>uGۧo/6d:$n|b>$Ep h|FiWyvG6biCUy#u IKy0cQ`r{,"GQ!ԓ(Oymh{{2/NIͮ6p'Y/o]YZGZc}ЦuhsyFP18IEVy}ҟiM[c-ڐ(=%LP5IHC`џR^x?Tu@"U@7i6(~tJ&6'Cha<{hK֭֙[7`r uI|ahL GazPu0ʠR׺aHeC#˥o ?0IDATTeE>f:&)$q HE2}vx=űJ6/G=2d1V;KRڤQpAㅾg-Vˆ*2 ℅hg,P?oqT9EG{6X;D|0$$kjѧZDpC$M})t4bHw=xtHS4j豴ی26!ȍϡ15IJCZeGռHv!}S iF%i'grP1fq}Qhl3K2L0̑N:ixxPf͒tK*.ܬdʢII+,@*ޙn%fJ[5k3e1w)zn'܃%u-Lg&*b4~6A#]볾~x@j7 sTbC¹XbЙ i+u\Ge\"Mi>e6 ;m]FUR>aֆiD96癴[4vF>3XJC#-3"@I*K4B\iY@qt+@@T 0y <9 $P`ەhƲ!]y6iBڅaE9D)i&>ɨ)X|3,n%Jb6~LdI.gg {<9QoAʇT*;9T# &DDŽ~u҇LI dXñĭGolxٕ?w:4z~V.,J~Ir) Tvj_i# ^}igkO4Bɧ1 BB1KPh VDsi!`!`Ҁ@TA,ЏbQ~Lub}$ Ù{ g)1UgH/CI:wx +s.K( 0qNBz0xБ: c+#VYFA"P7XWƄԧѧR_FGBX*rڱ!c{ 藔=<+,<Uopl%&\?USA!*Iit!`!`҂@T 5Cp$n0-2 Ԍx_Z:7aթ#ΰ*d2l]2"SijUfeY1K*,Fv%x2=W"h8%0JB2jk̆ :IkڮuHtO2.U-༜szP9 Y% A3 H fVDtmqݺuGJ0 C0 CXZHuF* q UOOP m`D9-YQTqkWn^T1( o$Q2:>Y40ɠ qSY=۟'+׆QAelInoC#L^0YTLnqTq2'X,~ `8 eUT$#ܙRK1u*L75-h!s&Ŀ"Xǜ Ia]nR hRks2~2S\8&6v9sEJ:B2ZB? #)`qU\$KMWֆM-zq0i,~$EIDAT攼Kk(CksA !dN Ṝ92"ni`!`!#PUza"~e d ORRkXJXęPŜFZ(0*$)\I^%TLꏘa6C*E Q,%LE~ii:y><'2Ҡ\&yL^D|1:,ӝNC>sv)՟d7[PJ:)u^Y0=#ΡNdi Th`&F]_ٍԹ/,OA#euR֙ZlL36G?hώ0RTiG=&1Q͔) RMhY`+%BL32B9Oō8d+퇌1B={gr>]m۶{>xC0 C0 C`iD(U9q~^+kQfHe]U~nBJΡ~Ǣ_UŹ3rt0$O?Y1 G%D7k1%EZ:); DmNƨS49ɤ C- 5 " vH8SȀ&y0k++uaQYUݔ!::MS<\u[u.xףz$ivB"iA0̛,Qf]`ؾ~4/vJ'=bi02>+X qe 򏷌 2 ]r9eG@mmtt%^yjqz(i}T]ˊ-0b;ڴ?@SNKc!`!`,͠rX_Fjg<\ @}.ɑU[/flE ܙr$=|\#+zN<h07vg$> SqͩJ0Gmi$urCo?w 4Ǒzf",IٴվS٩]叄:鲵H<qoc+cRτ0{ rppy%)1 ;SKre.C0 C0 CJ(4a~:g$ UnYYG ֛S^1xzSuʹ(r_^':!OOMðn*01[b 9w0|SϹ9=wt!Jl^IOl}d}6 mRs$0M^=#>\N9'=//.K@S d]*0X{a9DFŒ2nVϗ{|_sS'KH4,۹۹2wS[0ԟZ6I+7o~syJ }|{7QUg_M# md[q!@γx:e'\LUQ,#M;C79a}'j2δ̙3eH>qV#,볮:6I*o5փu.UV>\ rm+M`Z'26v>}FjlpR.\o~.kKc!`!`TWf0 >BqǂǽŹᵄʴ峢V ӃtM i#聴<fk;σ^[>ʜJ@P719'~:cWӷSiSG9T bN\ g^7Q^GXb/ً4S\Вa_aX]ƚơMO# Isx39p"߶2NWTڗӞq0`jqН2'8mѣCS0 C0 CXTb"/ Qdq) pRҨv)?wBI$vݛE}KWM9eU* `uQ\=b^nRt6SjR}X QFb)7/I6QGB[7"LoY\ʿKs QxhxwIL~I3 >?{h <_D]H~v3t퟽uIDATAc5#~`&#K8F^֔{wzzƍ2M0 C0 C.,ԩIwc_,uf1%)$L#,eWצ!C.`Q YG.]RV|s;ϻq_S,JMg/R[$~+x(Og\HHH39f,tIMuVʅ0FWG7I$qFHG-cGb? CgtWH('Ee7l"!gmIZq}sR;놰. b,)V,W,%ővNrS,%0t5 !)(bH^vR'?y)B9jzqW +ğghwӅ{2Nv!`!`K}%IuY[b'ӕ̓XɪqWsR݂H~J: zp`to #࿬-scܩg翘"IL6+udt^HquIDATa]/K= :u1ւP 609vҖI~Iҹϝ$N[Cawmj1/XHG}rh G@ڔF2}Қ0 C0 C,"up`6 j\\X2҃5#0O̩.YXe1~V>T7 &-{73ܐHE1sG>I# r!8__$iBy ̩T| c<#B;>R^VY!$JzE Kuv۴̩2vLxKydS[TaaN6RcIXZ|#_+{8Hsʥas/#0{Ft4Ki5 C0 C0;:- e`"2 8ͧbo8p8R彋%yM8p`mt k  f1ydTgLۜd)!Kiٲ.Q+rBM*1Ea3IøCVBfCB#HCBI&;nhHIxTD>!KHs{pcXQ(݋ C[;?\ RW >0,/ߚ<˸H mh6N.nO.&^s1imyI+|WG'NlU]rwvC¯-ufN0l޼y:*c\_TڶY~C0 C0 C!QDu J^dC_a$a9u*"22gWjUл),F[I,,IR>[Hp@u oJ]2>KZS]exUҪM~^yv; ө2TwЧ%O} 5բ}/ ^gzVi _WaǏ%r䧘zdq.C0 C0 CJ ~燬3RY07E|> P _iĤN2:5r=-uUw1xbxb2<+`:"DE3Wm/Oh^ߑtaYr`կ_jxҹI(h3S,DvIDATdƪ3;eXR}}3H'?C[pG0FO}&g3 S7V!`!`@1AxV?ǨcL`.6^ab_¤7#H ԰0\*R]Ѥxv֧ȇoObBҵp7G%p'$yQݒ|oʯgդ& `SEAz1arZ濞cxԴ<Úݻ/ׯ_i<պ5\1ano!`!D!> KT6F7 ٭pK/a1L[^Ycr{C&~_%*_OGnѣG 38tɷ.I-LFeŘӣxw|:Be4+0 C0 C0 TxVY,4!܉{O# fz)qx%7%oKIW?N)!KO){ܣH"ߝ۠K%<%UqCFk'/ʳo:tɨQF잆wu)K#CLSqa:Hp?F0Nj0/ - Me|Ȅxc˴̫C 9|})K4e.è:mڴhC0 C0 C, " 3iҤ?d ekӯ~%HhVs0oo8 Ǡ1@9nF;G8|H'fTrV#\j\lҮ]d]0:::\ߋ-\ou#ݕ\N]Ks){YI ./rۓr$zvoT3TV*DbJIY+iҼOϑuwgelS]J\ƍy!`!`!`,Y, *p a4@B76^$k ACSI#iA't$c<9ݛz5.}s<_-to⏏0i1LCq[AȣE{QAn+_0t?#>K'zԠ0&Tu7WaV4c;Z:\R;lWڴiS?ؐ܅vUe^'Ap]@]*ղ( ]#~z5!`!`!C`cPvZIbwٮjbI@Ak{o"#?w6jB͛7}򎁩¹urtdn!ʰ,K**ZO!ŏTi?~wi⪢ĊlO8 IsX}<w 6hB:?=4ty|Cۡ11I eK&wM9_ f,E) $Ϳ%r6מ:]GXq'--IDATe]!`!`!P T}`rh'E!7!!ٵA/7NHw#ϊGPχ0pM:uJ'pZAnJ&O|L{H ^~T90no_(M"6q>^4V!"u=O+O)+ ܏oKiSGjh ,5`V'\Fò?RUb#^c$'s+ jĈbfuF]ʒz0HJRxzfZlU2=> ]l6~aQWcu̽;~?cq!`!`!PuTUƉ'  /n1z2 ,AA03=~o'c ɸO"t|>gHSx"n~/%:^>/*}`CinP{+|$΢ S)vF]gY!NjB"(Ee`I*aRTyuvSeHi!>x]"{>1 I0jo- 蓃X1ͣ`B@wcRXr&PiAkݺu=BY S̨ʆ"yt7z-+'NK3 C0 C0 AT $}HR}-40)ׅ`0@:9I.'yG|?. _?MTlw<eJ T{eI\Hk! I@K`6$CLH&4JR=170Ϛ5KghHR].Izeiy.3q@C+HY#)4quըI@l6Kh5<[ڳԃG˹֎+)*ߓYBzOMRk.FĴ8SpQ ЧPչ^50}}7/&^nu]ݥHIO3IDATRÝNyϟH vI*x{AzFR= 5^OzI\a8[@]XeJ{ja@kpt#v@Cϥ9;~>3v,i177ɘ6 7#j!q_ɳ$Q@_IR-u1HuNRZD}` C0 C0 C ٝXET=NH˸,w1TPeY$<)$.|rQ"&wQΡ$1I 3Y%NuubFf4ar!#7!Vݬgǹ)24п:ix9eG e*>!3JЁ_]~=_~HG; Puh݌vb;6Ey8ӟ&smP4I\Z ,!`!`!!(AE$+S4I¯bH`ɹwXGI'<#u80c;l] wI?.B GH X1B+s 4}HŔq[ܣL.Yw>Mdŷ=47,iǓWhXGY]CgbiHI އ)`HI8jVgq.]_) y3jīJ%f>;K*zv=eok׮::[e͕{Ȑ!1[9N !DIMPߒNgwGR[I4Ri;MY/Txea9. YN(#_%!`!`K:L!Oz7 @'Oƽ eAשᾀEo'!?ex}?1=~$k;f_ ׼ZHkYy>GdtǨG*buHM P "_ZKcyZy$ at GB8H{{OIC5ĚIDAT$ ǤQkCBG?:ɧP5y07O[y>BI&ri@pHH\$rs1nQ0p&n@H)&Ab5F=zthbrIXf 6K>4<\<,!`!`!#P*,ȟ,p눤MF52 NHXtM1BW~Cy;_XZ.e7 յl!]s~K%xI?2kę0I2ׁ{2w%(ڑnEjj+@Q /O.GP+~1ΐ_ *%ZkӦM3g?YZ)TkS3 ,x7MC$X]z&Ln[pW_+rc5~. *N&^/ܧqtTgQ/L7@RU0-HϚL* $%ˈԓ&i_V;HORΏ;YO`ha-Ozv qmHp67C1$Xߒf/ae KQϧ=RTUR ޓĨKM"7VG|?仵#aU+0+WPEJCB8*3֖G!Hh}Yfͱ} Ť~ko!s, #Sh='?üݮ @oGkbٙ-%(n.G%;w֊dTcN޵Bcv-]fE9LIMvߖ%s7I8}+LqIXXS^xr eY/$q#~PH> 3/xwIdzysrE7'2+,}~̫XDYel:=*I,o i(ToyKl |+ xȪqS }cNl}G7ȳcpy2|p{^b/om6NvC2e/ZdFY_c4 @AڮIH%6_>fJXu? :. q3Sᘧ-P]C*UJL>RM \WShץI&-MiWG4(]+JoRF>T2ȊjӆX!EQCb@o[(:{y*sG~ͳ^V}U!Ns]:AɁ m kׁ)>(r5<eIٛ-oiK' 1B9D02\E| .IDATaOXWk^`$ ITai2wbKN$M,`"!xMZ:պW^d412/Pp S uGY}AXL C0%U4a:O < ]./S!s=6-ӧ.H4]A@m2dT^ٰl\mG*k.˰&`KdQ!>%wC0m͢V0\I]\ q ϡѡԤP)?PCgɂF3e-Zntt+- ?Z9ml j|gibgp_^z=Hݠ Hf/F/\h>,#Dp#!Ԅq>OUVC@c>A8mUYr%7jQ]eyOc_!䕋tѽ$?n3iS74 jH d jroJow6xJm@bB?k39qBhIg=oMS`4nmkqK#19$0ڢ3|OV{0h$)LCQؐs*pjы 'cf^I|^LtU9Xn}5^J`gWK)uv, K1c Q|?r@qtϣ9y:S*uAUcNE}gIhUl>lHF2]KY簔",WYQJ5;W@dįE;^0[)f~WnReJrbKw&嵄k8%_IDATK kP y:f[BZt$UFk)F1^TҲ~2/`?uec^g68[ tAaFwe`}~wRXlO?N::OYh9X }`KeJV;jg @˷lAULw -@y?v?{*jsLf1sΦI5kF%Wwg%ب1-${w7_<*tva4ee˖2(eruV&/%:"_֍>ߺ,+K DE5 J5AہT3KżU*,j3E3LjJjd6^N1o[MD΃&U *Xlbv1))G9]%b"g?\F bj |WB顬O>HDU;%H243#}2N ZtvOwrȶDR˛jHh7H{NDPĒӄפJo_;LL>L,6,)_ʋLD=GHNN#Y;vIhylC׹e]\G1֡Knh?`!VHzpԌ)[ucC6t-&QP?QTzLed4NT7'3r 0I狄BCc8@*`๎/І,eٌfWXsbCKvUD3nIFV%MZ[HW!Z<2LC].5Mi^ePv`RVn.wyXxoNiܲ)IISenZfĒc8mMIDATA@k/D!5՗2B@ؙ^_I0CVIIr^K] 4 kB>j4~bԲ$En3D4HW | !7q~DwA,Ů⤪uPиtDX)&;|\KrAo~6u2}ST乬ȮBJ$= OJy6e!|M=/r|Β- QY.9z$d .Iiv%͔t' W+*`.~x%} qIEDb1I}{Q"MX @^ϴ 4oea4+|U㌐%x~KW~M1꫎—z]ۥ#Yc"'ITT5܍5?tuj֌FSUU} $k2fW! |{$ӞT6ZYOfNwlUW:k] ;97 F=˷+1 ܺ;L FW* +PN.dR]I`rؐ V=HYԹ[`N~=Yz!K"RF170Q HO Yt-R*:K3{!܌!k3 裱YՎjn@Vs⊧6R)s=Jmwk|x;1HĤ)Lʇþ1#IDAT\ 3q%ɨ$. H%v4W2T;Cd0J'&O~:bu=DPkqCGY2%A;۹NڌhouԤNb$R-֎tm/Rth8ZigSڥK!|J3B@"mwp8o9 {4]C!<Z/O(S֞emKƏHqI. g}r3sD(ˊ/vIËy|~.+xyOVK|97@rEcKnz.Y^3GOC䎪e.#-1Ӟj/Cx5t uESBR[Jkb,g(x5i6 X}g_7o.ԣw:>ͭwA it4ȶo⒬B0꾙wh>o \}=ۮct}QDŽt\rrξhkO`$uq|I{#4,/3}|I>.^#ۗ6⽊>V۳#ijխލ,Yu"N\Kwnއ[||. s_lٳZBHNx'/TOxI~wZut$cB^-Es\)#hC*+RFkm%ϴ'432X4.-<7@]كjw%Uuy:|$42"5 :G{ L8SD"IĉcҸq!GQܹHU8h1nHI'Lz}X#'|Ϫ &tn]9S:PW0wY ,8IL\n<I>ɴڌxCz5<g/tﵯZ@?>-l9|\iB(BXz$Ρ:'DCĹ {^)uiB %~EXS,6?h:6KҋkhSA7_z/i)#s57QsnΉZv9&."L{% ZD1;HQ/Mm0fۗdƅ2Чu2$ }>(iԻOS>▣ EywEƺoPˠj̆4O3aT ;bEb 2@Z-1wO#gpt&Ieɘe٣Yomk,i=ΜsZR?͑^yM=mh=k l5T_#UJd `t[g3/ ҷ4_ΦП/"%|N3&iePկlBͤ ~w;g8d|7nJ+D<+dž_C[֑;@ڍ#c8\գh4?s(WWIT;}-uvLniΣ}5O2|uRT4/wՑiS-n@^ :zy,! @M8,cT'mqQW9L( 6 Pi[i,HuA]P{ܚZnG-EIDAT ٫qk qBnj,LM47EMr%M#ed,mT hi[QՄΛ wsEכܿ[b60#*w%h%9dPR Xo4>д{х1ʲ("Eİ-(8E۠ɓŶ6ڱˮr@mYv1*{s 1bx^&\4-tNthcu̙#ٙ3JZ~Eq2 8s*˹g h$.4'1JVSgћx6td-^I5fp`u@H8K~a8YM1* ,{i6thƿl3ܭoW$$K% nPm$kv$K105C)9':vH@uG:so}-$~˥]2?]<>!*+Amd1$#$ʸrpo$LsQ% 4DEi2$ Be8\RC$H[^i32v F@jIڐFJɑ (ɡ%Xw. le&CAW4x?W"a e.W([xmQRmNV|84Gd]; 4 (S1]lj:[⣁vL}`i߾7 ;v'm;yIp=.U,iщ`@4ŠhwZs}$ $kaAJ1=Yebq%vp{MM8h6Zŀg{*CXƬp>fFhԨeZg቗h0&J"=Mۦ0T- Mї6?Th?Bqټ ?)m0Х4ҔI}Co!0}_R΅t:+483mlqR7Z dI!1aAHCp7-ms >ZhZ%!&4e(GG9TIDATYǘNNLks&^H:yXd/Pk^`nzWA Em괱ɼ/@ $AߓpG{0Y&"+si6<2H]e\ {Й% Xgh 3C],tښs@xs?$吱YfF{vbT5q]ð\-ׯ9&zΨѝn}BZhGbS3Ƅ-eL1i,DbfG^byF@{ԥy#RRL٤^Dii벴Y 2^Ű1{ž?2 yIDATc)kt0"7oFҸc/LQu Rfs6c6Ë˴\$zL`$k~"쫱@|* *s`EDP+?0 ;̅8@N-&rS\:65 L4^R8Q^mIS ́#yMx9R,/:NADk'qFq36^/Sđ>sYR_ Ga#Ș)xJ=4%$S+PӬMK#] wU4 Zp'iloA9 d !MH[o*y_gzV@x2 *SĨ5bdIRIYHzH]ZwV)(zqxўݣ u PMbz9&Pf䟉46k|>sV*Ks7њ5>RǮĤ>0zߤ¨:xxt9,.kvϺc"30*Zֹlj̸Rї6(S8Ybf']֢M*hLˡF zAqЩ0)FJڜ+㳠1r-RR{(s :F$`YkfP m$%=PqU#PuL J j^h>MԙK %/9W,x:=>49sVRfbV aB*?6$2Q v X:)lh ZCqJq[K_=gA)ˠYnO*>R. 96Is8$RST'Y/鲨30K5Oq{1.~$EXD$C{ʓxֹL_sK.i`P Z;"Wm&IB1Ru2822; @%.wDEԩCWnIyI+IDATdGm\ *y$DseDiB܆A79-RnM;jrRn"헕ぴ" 4k۶sG-CrA˞xK1RݹD}.&YZ}R٤3쉋LƇ \&>Dϓ 4nG6hC`=ͨDyx1/t & NIiy%΄5&=wOjOںwߝQ5.J)D\Uso8T'mm璦L鼷sU*$m4=x:[w%qi2GqDbM5SbҢSm9عT00OJڛQG c,7(3ٜ./(ƅ]퍪p,[:ks^ZioiS`6>gQ%eSum%蔰 ]r.pN=wB>wUK+g# 1`[RJ1W$I}5@0RzWH6}-=l%RUA tꂬsgI")b*7" +Y+0U?ĜUVbfI=g~ R-g)C_gadY% L1KVz >Ԕ=!1Q(-Y xg!eʜ#fEF-#DRM3}wP3yWωePk2Ism(TR5EYAێWI~'bn65T a$^@?^*YGrTX{k]WcX ${;CF hwl6}dؕZ>^O1ڦw2|w'QR_i,n`9]IIiį}UlHCB\Ʃϱg>IcJSl)fDǎSc6\Y>z0utI>; `ub2%vezjIDAT@w_\5Ԥ[;A /uc3$+5˜=,l=M9IGlrpݸveƍu˷ E@bfW+5 &305xrphC9bȸLKKƉ,T!ȀP`T^BPӠr2,U'sPH2^,JOR&R@=G{ AW4s=ն4 lw`&m噎.Ks*HTM2%dSIm6;uU\I[6;{b"OԩV`"iˬ*ʮe{Rcg}jB"'k7 -}W0/kj@wgT+gy,JXѭ>1i # $"1v\v:MtmLNgjϙv=k7 Q0~'UyMyYp)oa| p.dv-àb'~o;]ԓZރ3KC; Օ}`ʐ;.02pה;GqsIeIDATiS=+HF'_39=t$2+W;wgRGS5%;WW23bKȜӭ2JImR ܧ 3I憞*d (^|rEÚP-pɈ/.w[`AM\j8$ ت+i#qdǻH%oWzʃ6z'ʃJ(3ywH$5i7ӲsTI3 -mˈu/_ 8S 6Ӭ&#h+L[TM,/8k"r0zо|* ~NLN5Sr_!F$ 7 4,K>Tk?Y&F7y&i>srrMۣ3=&!Te^@=0s(w6u,S9>J >4ȇۍstu乬E1C}hH %NmBٝjN ĤوOze$Rv!ϲN%VNu꜏?K~8J:'# 2ᮂ }@Ԃ>̓S[B. $SqN3nVb#4}?59JnXIRJRO}ά2I?dȑszRE /U@ۃ^ϕ1^ Z?sW؜XLW(wIjqL*=Q7n< 1Cq{&eIC%y4CHL2?$)1|oʋʗgKf\^Ud𭭖׼vA!AfP ڈ\0K)fl|rUh%&U Q 4zwnxz0XuNDw_Rhq+g+6ʑOHTKOs`vCt &JK?" zI܋eF|$ MzOmsRMՄ`GT"He1<9P ph - [,_")3s[ibIDAT so3/K.*$"Š-G&@̍;E5fSj"RyZk5^BA`z>EmW ]n%Y` }1&uwt6f1 geEg®ʯbxѫ.Yh/9nWwƘSQvYl}XOPY=4\GPKmd[@ 11 F~n ԄFtM@_G5Ę ˼xo@ƳH&2RiB6 ӖaOB id M4sxvt G3\DAy3Agm vGoɳ5éoS$u){|P܇jtOxLL+unaϵ9@8LP'0|tżES\:OVu~ӤAquEЮ-a֏rm%37-CR0QW/gB^#EYƊnK"i*K:'7R*l0ef-%K_VK\%˳;G޳K79H3 n"xܹL&9{բK@{1 |sSI<sʚpX7H]H%.8w&8O=+i] lfމ&a<㴋&hf s6lYD˓_o&  {ϪVi/pUZ:S >tCIT ~0/]jԨ90M'ZE5cf}΋6f1~>82&$@d ݤ[Oڮ0x? Fק=,I]::nQ IDAT6-)umN|Vo(=B߉R萔vބ3@kؗt:hA&仠$^}!ݒdn2N ԩCﾬ ũ4f8؈!c(nˬE'nz`LB9 LHk8ɉv|xtyC2lUOsenX&:YlK'??RT(M1fp[&IRBF ?.vKQ|Ksrʟ8q =Ws9UYHIi2XnQ=rnQh| ;o@w8Ųwh\Z,]Y]L9dMϼ F6]Beeq᧍(֋99}'9uL$)jU檂m:Wͼay˸-q^B*O^vbfvHDceW QuQY%ds,tIÑ}#vLcL)\ٍ :0,Jڙ,CI:n%ƏKј9'cV!o_#H, T!;hC^"`"q^fbZc6mfC 2tq2@dtn'J仅r^p| ,yM4QztW)4]@qw@'O4xNXll:jԨ5RI/ƪҝ˔˙mdjJ?PbYl>0͗ײȰ[:>+1䣣ВЕٲ<iSy1#+ x5.]d$lUc|<:I󛔟sV-mg$~{= "]xZ_5yb7tMٿğ,iof]MUxc [C8Ӻe_T?9C\K[dF Apߍ6Oiqs47ֹ ?@{ISҜ 8N0"%ґ$X7x5Ǧ4ոյ唟<{Ke7 K-iGh!^>*286963흶ąw<[FQ`Bѹ!ËYy1r1t6~%bpVZz=Xs!uc OS7sг*&1"P=M=gy>[\hm{hSn֖soԻWEWk[_yXAHUl_N9$86@uIka][b\-6DWX=*R!W: &o_j u1;Rh`noc]zo⻴y$D1N%lMpڱX:j1ԋѱ"$;B%97z'.ҽ79 _"O$VHi6'EXgf|d6Xj)(o._&׉r}d8Iƈ$q'-R=ZK>i+jRKZfTuu>TR)p>Rj=Zd)q >E&@ Qgu_z[f] /MO*;1 `=gq'3ut'̧9/NʫAڿ'狙tZY$6W^>'Zq,A. c>9:ﰿL׾U15mդ؆>IٴQޏՂͨ5ٳ&;ԭ(Gk尨5y_췡.b6g<"WLj>M˼8RZ Z.m~iS&mR_599S} 0@IOBXb$fD7B/ 9&ec*U$ ?Z@wf WT,K"uYɨhIJ>74OsC'(}$ TQx~t /%dY-PQmNR JR&_2R3$R'+jb{0Ӗh ̻F=t0OS@7]a4NVy|0LŽ#7q?TDw'HECdY|"=igrU ]V' ZHM<Ei @Qy@;Rj %Z^<cv#O蹝xʹ+%HgIp8rӗ5}^dWVXhO8IDAT^m,#pvmh:@Keԑ(|c:UbfLgIm-st c'Ҋ6,/sJK.܍ӎ+ E7 .}$ݱ'(9̿{wOҪ{}%~fΆ΁SUm/i$Ԡ]kwisbǸڮП{!&I>ٳ+<P$Fw#T Oᐐncޑ#B;#U':ItC<í$Iji#"*֦3md1ǒ0~, Qzz{$M=d:V*]ڥz`P-F{-˹5  .gtRro-\eA0&NܴӬiԩ/uқ{$[MPҹcE0''OJb Xg%Ӓ胜%a8%Ew[Fi'b!*K3Zrw'"#2XHZ ,HG6i'& ݹ4jb2Zh05%ѕTf&܃{3~>c"u4g Ũ=*}F,t fHu=x<=# ҇Ai~G;Y /#u;wIDATHCݝ6hCs1p]t!xƁ6 9H%~⮴"74nUePx-i3#muy61 \P@IꬒD?ygoel_ܿ*m1/`ҾԗIko;=Hgwr0>^O)wTG( 'חX^z7hӢ!~Fӫw`v4IiN+uיHr];i3uA~?*I3IE{Gi2;%'9{'IӜҷx1y/vh,CtLKdr'"fB* 'O4EIsQq! z^sPp$`I ~/'x% .H0ޚ#OHiNs`(%mh9X͵E/_i۽èW}ˊ=ݖZ$Ӽc./C-ׁKgn&ntA?Й9>wy=gnD|t =7]) )8x*M,zpU{2)nE:Qө8^}yI+f_I:[lI*8ɰ]}obs6.eahB{;KG lE)9?{ ŒU&ZBT&HeP#8';#u7s /hP. {! x`a>HJ^Ǥ;i琒ǒ)[IXXnLHF)ga),҆2h}t:e`QiYAe?K/Vjs_"k)TKȢNJ+z6􊩘]0mAnsr4&MƠIDAT5eX.$DTΙYps=xo%a:$,=K4~gTsIzz!+$m*æ(-TZz>mY%B% C0 #M:pe(e-qaKI4  Hf,ImYΠVrֵD?Mɚ+i kgy7gq{ SBݒ7Rp14洪e!`TmduώȧRk-2J?*ˍX()e2E9G"5Go+ȵ,H8UBY0BBQ]:Q#/&{\<{&UwM(Rb'kPl!`RjڔÚRyhX*mB{5&̺7'0@/ ð{sVY_`>>/LL):')"%5<\L; }/:cze!ʒ[u1F9th^ oe}^>+0 C0 rA@d5Clz1jS.tXkd,j |*YBPÏ%vcP˹?2 =  0ֹ{b!%jcB[¾f?Ҳ&%=ζĕ=S<//ELm`Rn@ooa$;DB8]I#rONQQIY{/y'#ɔOn=_o埊:/V$~(Q7Ϝ9s?#7 1+(ƪ19W)$Km!$! #zt$dƩ dIDAT"Zgm']Y^xyx+mWNi|WŮZh2+إ2ϾB/y6.DH:K"Cg͚Fzdyo|`M{&g{W/keEѧvju1{P,.I52n0s,9fÆ `yfMrP̋ךr/oL%&Cp-RK*+hlMq*IZC0 C0 F@F,,vUAA"vZsԨQ*| y>Qrr>8јċݐ{=l$Iݟ:o|&Ne^N;ƍT'O,(;V0Ǡ&,f C0 C0 C(5Ơ) Y!?h]=z佇k?ISNt,JЮC-Ӗsԩ„ jOEƻM۟ޚ{͊P]к,Lvܟ'm#@Y1k]N3_&"ƷBB[OO=Fſ%tD=}wzӧZ4^s5~,4iwfPhV3 %;KN%!$"0f̘0zSzu]IfJ|?{ըQ#q©8sCK(!O 9U^Ϣ^GCMU9hP?5jWpն:a͞nݺE#&5`>۝4³q<͂m_~eZWP$U￟믿'30U!F!`DRh Cj! 2߂j|_HG#0_y,GѤ4Ta9iG%RIZL8Rqڀ뽂C-|^*Ub,66j(n_z$Gb26˗,JrY`ce҇1K_[ IDAT˪,AѣG!cs7'B][n-ǂTIARTc f,ˬ*)\ԥ+ePbːI)1V9gy_e }0-{7tx:z@@jMF!`er!y $1yIJޏ8{:sHWiWvBj˕M_v+:O'Q-)w9כ)saƟA4uvҥx_IZSQ}U5ꔦM0 8Ơژ0 C,LJzb!/cIIױ+GsbԻOS1*W) eΔrTV7N[T!}YߝѨ63 C: ` ju%0*ڵk?B:I;R5  lҤI|:::U*V!PV:#,VeUOU+GG p?mLAw&vTEgI=Q, `2B,bdSEm’DYٔA5@Hb0ν}/R}{SN:U*ړѺ8#f\A|c;MGvA"FAr}+{!τf],f)@7%B2zx-%8h/Yu܀ӊp3 *[Ibn%hjfӟO۷!pϚ{_Fy̓gqʯ#И\xk|?M ;@G]|;9@! ) (iիW"Y\>|>ni?GE)yJ!ʟR*@v,T=,^3Bp,\VNak88952 ) /m1k7ۭ" ʢfC%mB/kk{7G@'+,=9tAwe\ܔ99w,JZLAgd+U~B?Yؔ{MsMy. 4{'wgޝ(OCT/Z?h\ }}>%1}76%|} ,XB``Z?XPoz= ,̵doq%^(7D0>.7e}3(nS0*}ʗƱ֨7ㅵç#%)ދ/Xrȵ3E9 jC 9 WP;x9@jTå%Uu |F=~eCFDNׅT"eJ+6H_\5 ޽{K(qo?37&2]|ʅǫҤ#(sK9TKXVt@-2o uwR\E\lHIK] (?*bw{,ŌAӖ>mBɲH{\GaٽQ^B{ScA:`VO-OIDATX ȷ}mb(!z+DkڗuU.ASڽl>F1˗ du*sX悍ZԉRC4<%])4ffvEs1<5ޟY(Mů]1>h/>e-'ҋ,Y5YL` %79P5GӼ 7F=go<4ʴ?/y\庂Zk#p whCsV3jԨDى/5B^UʛAW$$bxX:` dɱ !\]Bp)(I㓨T7Q,\!S83B8^,K ` e^FwF9Uݑٟ&aw=iIr ̞_ (4.ur)'TWu2c-TCs }#_zꢭ@9vTt갞xȽ6s`՝ȅ_b<[ o/Q < O{SAҹ0 aY\)={%}.\hZ #^#`HksV*ֹ 쁞K_p0s:c~ %IZc,KhԶL /t\A$l:@a6ʕ p o<=IGаwgʪȵB][Gx:IteHW "}w ʽw\3}.kT\z ?XBH0C\PHnddMSQY!rjQWk{-6j,}w^2d -¦u0(p,蹲p7˱>e*#0Ŗ .(f5~|1ňBGSj|^+QSFLIDATD,rWoڬ=Y$3]Pzw}g,{n]_8`|#n<]+S8WP;z%=t̷bR5K-M2l*K]wtj=KAyVTb8 |WSYu+]δQ~S3E] Iibt+, /C.1m['J3A^}uέ|4kcVzÀ%ܻ[oN8WP;9@ 039s7+D+K QDkAhkI("5&TRmVH(mҕI5+jE]@^Ɋ%r&-TYq \wPCQKVd$ >r/}=BVK+vP緂yyѱPQv;?%QgQuɻ B@<k,M=PgT'/ŻE_cнb)\؆K(}OjKQKڿ ggԙǹlm L&} +\Ccb1e} JQo+4Ġr@3gpj"x[(ڃ@pOW,ԟg|_m­zgw6> /z-6ffs+g,GE@BD}Q9"P9 @|Rڝ]ӈRXM LꎥE2^oZ}]I|J8 W徹!kH}߯$`! dyόq̿몬$qYE%ϳcݔknUZ]˖- Um˜DUJv9d]Y kudMֲH^K@ѯ؇ϺGYg-g+o}wC݊ W<읞p^p:4vq־Q\Fa^Ni7l#t(.hel |Y>,YWiiEQfC{.M)*oK5)n|-v?ܔV*O4 ;gg 33]>ڣ8Jb y:)+|ZҀT!TRQnq$"3/ }εO4q ,E/[9wp9c=&T'2{E= >}7*Cϟ65GS j&ghr@T%ʃZAyDTL%}8Zk(H>Je J$jqp5, Әwp?m{ӌ~Ȣk*BU_xn ~)j)_^UsjuV-ZzU<;')Oڟ_?V}ލ}eeu27n,J",<rdBNء;,ؘo>*j6 ?)M=p* ZIy;nHq5WS``(f }QxQvJ@_8(ilײao0邲d#;ӓ@+л!H! sCR(t`!G! jgXYDJ!L>IDAT <[qYn\RJ4N^b˥QG]BJBU!s<=(GZq:JE20IsN[K[d1ŭA{}xFƄM6h ]klW`g)U-oD eVLǤ }K-µf^1ïxd> !+卂< VYt6? n)`eOx)ݾ,+pKߛiH7@bdgXzӿ̭B~O6_5wtFKKK(KRDJ=uRf)*OsNRwI\@8yw",* B,+zu\K|Q|/~𮽛~WRkhh0I.klgDQnvy!x>^)9v.KuIDAT?@@\_1K)G~+uM+m%ewL-?ox ~(2Ө{0;< &Q>zSW"xBʜi `ٗ ݿ}-̳5%riP-x?X|R~Ѧ-F.Z辴Gtp ۚ*(<#t\A(#|8+b|4PR)_VH)sx.9CDiDhA:cذarG"Ԩ! {.Iֱ?^e1kB\ {?UST UWCn(0.D?oFQ;J89iks-)=Ì2sce-_J;;h,Yқ)-gk_-!߃V\ͤd(mA\R2,}(YaI4s0< Kcճ2G3.jR |,,}4|RpܻO^Pj"YH㓪CV~+=vo-= N@ C˗/]y^ nJqShOgLKqS؋ɂ w^{7|>fm-'8,ʳt:=/˴7<ϲLWxXuG:u=q!p'Y@G it=B{kDpS=X5K$`E#OMI"ɴjN(%^+?q/\ϧTZU9!ʢ;iҤ9>Tk}~)Yp-9k,Qf$T%[(38h_k@4WWQ VRxIDAT=)Gv=.E=%p"QHD)/!qQ%@k6|.n*4.;A]GɕWJ5qM%aORسq] 3 Kb9"-"oo긝sV2,댕lY,˽r[be&UyWiпV~Q k~ʡt4\~~!tɓ'3.;}u1N ho>yQ, yw]N`z߉r-n+3n\/Q)'~,XPȧ+Řr;jzz}9-4DCsRZ |w!pe虩՞wGpN@YWWR`3 NeI@`eCǬ!]|*Jr9ҤE08|hebўmi_NK|Jg [().Ya Lğn^2uuT-,Ik~ˎwC,)[wLw@4Vzi> stream xK$Ҟ~8VKy^3zD/5>'n4ڝt?__w{{㿽o6-/_ܺ[?~{_>?|~??|__wYo>w3o~B/۟_~y{v?ǟ׏/m?>?߷Zo?o5_Xşv'J`N~eJ/3ے}kR??_M;L-7m3El6˟Os~y_gzUNϫxۻi·æZ Ϗ{K'dp1L2>~fgߞ6}|/ _ޟێmu}A>_/x!װPBݙxͰ 6=i5e,ҿ~a432kIpR½*v[ݟ,+M:">G>og>Wj|D Qn󌀎 Nc .Yd}{)z/|1kdՠ+.4A,~~!Dv[G~5iӪPFF nO}-3 n7 -4ApBs(^8[w+c=58H&ۺdze[fﯣqGCz[ۭmk$40qI2Zuf#fd|gyz8q3+]#4x~kՏFfެ?~럗d)?k~Ő'A;ou'2l*Bk4|2"%?F6os{K|3LzAGwm_yO@{q_~.^mę3؍?o2*i"i1>)cCz^8y5H[} ZAžpMCH%"y6SfFL՛r կ5f>R붎!Jy]J8GKmm;|3o=1 uCLjC+~Ɍ<1?l BUMHlDblV6rۘùi1^D-ni 4.K0/^~ sꟶ{Y]ܥ,j#Q7,e_Y#m$BO 2|ɴi}!)c3*IcMdLP?}'Ag~@4UJ..]iK*0H6ﱖR&aw76˪| ޛ?&qe _"IێN'b֤ÍLO׏O8̢}&3"LOY% -e'G>DBrJ?w!)]6O$oC't+u,vL^Dm-MO|DB?agD0u]J~F{͢33GG- 'tu=f$YT2_lFÈ}-7^y |Q 7# ngdїϟ_03U%3L2, .jxY"BCZjiBޜ4ITwy`*)w3VS˰/̔X^錇 Rۢwe__s'.e嬧Yzq?3i)Pb݆eQʈk }ƃVؓy,Ja(b_!f3*=Hǎ^g$ x3qf$:szO2g(*&rIH2!D0\F_ô%8dЊz\9DLhSt$߷Ik'pZpeG; ӻb/~˦{W6fa>h+e< Y! s^Sj ƖQBc;2(ˬ'gȓ)Z[*S[?8U;N|aYzc3~,gԿ~?(e8~2~V$T MpWA<6J~)(CXp q6ZX9grHY3)n| YھOi!iC)ڋx+!. :hrX*]ȶ04[g"Yl,ؿ%De&A 3Ybp|_l`qNf÷ay k~O"vİWߎVX9wG ')]@AxTMM"ƤA|nJ gyf<sq{K/|-~,*AW_b 9_CD;͉.,=>.ߘS9ϩB.MMO{ 3eq [CB z9"t\z &Y} ;ugC Wj-[evRȸ2Z; ^1?,q4Zq%<"뀿dd~46l3Hj75g:&-Al$zӋ o bX214l".?TP>b/PFc"E8v.\cfk+!oi,cL:1}.c@4U w?z_C9(V7X ͹̛Iwa64&4~ǽB PzCst'orsa=WB3h!⠖-ŽX"ϖt+l=ewV놴_gs,n⭶U,o wA,=SlD#ҦӅ{i/o,=fuDCwR(=+&}\=(S1@W=-h]!i{gn@]0grlv7nnt0>2 k=Qqc .Jx$l<,YRNK7~K/l0gN>rtRɆD늧[^XY߫\u6.XS+~#j=?w7+W R*4+n..S'qsp \aj?52,ˆwo\ȉPVg TPpRp.d6i1ɺÇO_c}fn<4YVhtSvЉ\s!&,9_ZͱPD}.(R]&bj퐞mUNP&Ξ'wCmjhwᄚEP 9 HqmNz՞,6Ҟem q ӻȳ6IJnS)ͅ@"g ~!`ǟ—iґW"p."fԥnG%Gl cfٙ.` !Fۄo2!V9CK3a$.]ٔMJd6 z/ǚ2b<]5^ٌ$\s.̐m~kŒWv Yl{ڳ=Hf^OKm]yO+ĥ9\]#($!YhxLN n՞T-]^~Z6hYdݿǮKO}$(lLB/4YI=*p̐(lpl:ͧ#yfρUXZkt+.HO_p3Id517KUb r9 W~MiV@6q4E l.|U}@7Dv 6ݝ)EnB v"];g9fSKHVw,Ř1&#֫={=w;PT{`Sglwa`1)R3ub] |暳mӊ)-L k/J<Xpvkr*CLVSi-/&Bǚ:Zg_hn1d;$ k]xM`)W4Ks# uW(\5sbe͈%ˏiIoŀQs= d⺉v!KdU_~|T5-be`@V;'W^-\DVp Zk;Y۽.owI1AdZ\r!%-PͽLBbe^ڳu2՛!mygݻ&Zh4}qZcP{fe0e(%/G9P6hIتS~o9Oυ(i4LKH0ڝ~ ](FR>+]RgpbӁiǺsLdP'_YiBVp+ --/{ċTFE;&E"8 EoQkFߦ `ՎڿJ[6b0nF7 ^g0?0vgϚU+lser^v!k`cM(,;qh4+r<УZ!w!vӚugX(}({՛78ҫ{y7*X-"ہ3ƙuN8#d _*]A{ gD ق':f\!4mi6s4rM=ZGLk|4Xֹ}g&U.6lr&NпU'K/fUX# |6Sop{O u~m{* ʇgt[9)Xi|G/cL4V 11G'Ng  ݳGV_-I&$ǥo7i!^~Ļez.Kg`;ݝ?S(+x=zT$$s3,~kJ&<{rޑu>(u. }L[”<;S6AwXSP\LBϴE]wDO%Ona&;~]{O]H:OPx`}w<]O`x„..\rˍi]nIXo,9y'V'YQ]AY 9ͺógq]N_&׮*ce ?Ucxv,Jn\ Ý{hP2.~;7+8غ?f|"@1Ι5&i79^b: %m$w í'<'ӞB3#} ڑ2Pm'RwJZEA..gMo GMfz˱\p 圈qD0҆D~ASqA `T핬^jmvpk&ٺo'}r2raV7y#J|nuk흜xO9ʺ nebUp7]V-:^u{(O;KB3R~D` \ra T'DބѹO[!yZ=~Qg}K'3sWiF| vsF{E]N1$FMo 龠mƪ =\G, ~E kˁCo8ۉZ,#,#BxI7CDAQKaIYS a8&3qGhL~'{;eD]YzwÐ^r0ߗ 2?wGL y>Ұ뒻սܴU8G S{G;30/A_6[Z{_B u}ae#JCٗZڝ szI#lЖ)'b-]lm3_S^L%ǗME/?4P {x*qJ{UB2݋-cJz֪"lymMBnq{JF XwRGG6"|2h>`n׹& 츶 oq_3:C~ڤxp-c SC4\*ç=.G7tOFOmEGwQ8D~'g]{< 3 GG>xoGCwxՍΚq?Z3TӉA|H塴!{x\ Vп%p>YN,cƙ@96>Qj%q]zMAH~؆7]Ty@zqh܌Py rƎ0Ȥw_B[uo@>p}Ƽ>S"u?8ow߇EMrn ø +\T xSVxy30.–^.+DIGd+r +<:o`YVoxJS*uǯ _^C{w:v0E^L2ncYB5/L8 *B7& GA z{nmnQv?w4[+wr=wwnQO%xYy5fwyql&.:;,!O& O6+]\cnL == e w4Pќq H͸ӽ5q݉L N3Ri! Tl$ej"ysTÅ5uN0 iv2ECx{+5/=9LD[7އmq~/ 8o`\'dUkm\K x/zdiqQeS4Sc£LHy7,6o1tmVe+L><7e{|^@_ϙ&g8eXk|cI[˧ۑ}j#yۧy1q kN.“_'ý;`~%VW:;5y繭pCd7:>y88 $pGr @Ε_P}b 3v-V-Am\2i!mȸ5[^aIH?PA:Kȭ \i;I|Xyj%o9$r/4Ew6 0Q !ŮQ0 XOvn8%u>9F;B C;&m۰ƍk5洌Cs^]9ܔ7wCd+(d7~D0yP[ Bex=d󳒽bIPAꮆӊ"k}UBs4Bm#M1ܢ, fKEmzyZ5B.m8NߜBS'"zZ;)AmY/ھ ?Xk%a-^PbSl!h j8M|{s~(I)0fY3LXt~{FčB-[ q Ȯ{Z; p.f,D2sa :%B̬;EbZEI BgLi!њYnkz\r xOcaj=;]=Sk\`&Xjb/BCl~ hQ9xn4!8ݹmsfǛ%OOHTF x*= CW݌΢Y$[~:ㆦoY-smK׃|7z~=o'˱&W~;dۋPeEpB9+:![ECIӵVoHYOA;ۑjf6#c<4ܷ3lKf;詔c3TxR W9ݑ<ɥX30s o׸-(qZb) q,\?dn9mo)8P@1tٜsEbE??\|ݡ "%34, I$P&Ւ4[J[`!{c0]ƀT!<!"/!;ҿ⌦-ȟ *wt :X0;@Zޏ1wVV-gO# [$}6z hw4߲ EgL9ef ЎPXavDF_㉡qMLHhϰT6w&V[^M1mvmDIRJQQEA,=(Y!gNc'ۦ9}=#c8gh\"Ek4lsq2-~,;wev ӄo5nhY\7# mG |qwq6y}=ˋp-]L_izh8ӍtNJN"&NCwTĖBKDy:Vأ""ae%wP_G[n+#Y*Tk,&,IF> [~Cf6 ;ڗ$&`b/LtD]Yvd21.W+O:buJW <1Hq?k!),Y̍`5hSaob_"`t/maqp3X)"<dW\%!o+y/c%c|+ "R~mlqp-vz0@T٠-:2x'6Ή4olrT=V !{c_:ϱ"zfKYF,  yrV Qwth5M;1Lסz6FiZ:`:<ܵzɸ?Y( ȍY_cH<[-`<:PF5RR4"m'h5- vEi"31rYJ~4rHM|l>kF*N°ڻf°::V<i,//E]|Ͱ{M.>wTZT~/LP篦N ]wNYTY'}=8f ;a41W0B; Ą@4,V hd,a3 Զ_fMz m>5&O)WxԾ…a4XZ(i2RIwC3aFl1}U b9M6%Hs ZpNu0bGA<ÍF&7>"} Z W)DPj7N9/裱sⰮh7 %Z=&1=j%ѾMGJm:L`k,t/"_+Wq^c)Ja){zфU0zak{odXwY3M~||t}nkoڷ4{z .:p ^WHy/ܭ!鉊5gyrfw| tԌmSzdqؓTgzfIS,sqZajj\PccDZ,T]a@cjKe1 6)L!J/ZJ2Paᑴ-> yBDDD@#2fW>^9#SViKmGڤYC>k:vl=7IY節 xJ?s"@3zoʈ2Ej䝴dLrG+71'o f}D8cAϝ?jTGGSzJ{t4$;|jp Ml|e,s"trηN]yKqifH,A'w!ŘN-ت('P 6Ģ Y8CV/$էf9B^uͻ:Ll*6cQ=; l>س9N7b> ρ.%^+Y0Z42 iICkXolt-}hxJˇNuӷ;lzIwCw94P ԗEA*$]t(}6OW'Z( T>"'>Js`!< >tDUBA[,̮Oc˺!Bs!O6ai]U[ktp3܅d#GO)Sr)G |J1RZkS:G2ejih LI8&)׀hc_ŗT'3x ۵u\IH. ⦫"'%t05;b,)D;[QjH-eYlyR4Y3Ͽ(vq^~s|rlnuwWf9p,`|.|XLrLѓd$7ܞ551G')`ڌ6 yɅ&*LcY36"PߖnjFؔoM'5$Ƶ`ʚhuX|l\im#dז'IC`CRƏ,Lk{2˔~ü 4i|İ;sFLjft 34yz٦XAOiR3MHY箊ifS  AB){&ƞ|+q6;A+׽U9LRȝxSdq5מP&񘸁@Y)-&q"x _:OK--r#nfh9TozSOʶPت#2[#Iz8KM c%MF9 kBc1L߉`K|lͰ=~Z,׸L$1/1S~̓s")M:HDldsf}jEcgDU{}5gcfΆcFs7>;v{;GȃD#ܧΌ q| A}r̊u W촎uV)8Zo4`l6^$ ;uA,}he*^?2 *w_p 37U9 (t'Sj !gU.fAo=C>ܰ}艳٫̄45$2φk$ؽ\I)"]DS&˪64iZ1+7?ػՌ<, @9B+9/2H ? 3ĭ,\ uxk/ ?R`E,$<&|bc_F4l*d]?!bGo[`T Y{IŶeZjvdؑFVLL_ih(k ڔ'd)*հ>+*8=1Wxʤֱ_EY=bxt= 'mXQҕsyAhULAo{?gr4V jFR1Ω9c)FfhX㷿aJk}?Zޕ[$Œ+惴Jb"e޶f\m);eο{V| `چt"" ~v&Ɲl~;YyY}#3.ϓsr5P2Î-X%,K"|C9R)eHPyfk5viԚ|J{4V9,Gyڅb nNR3η{|.P. s9qN[+7D0[4w8(R!ˀ*Ӵi\w4o,cbðEVDѠ^XFj1ϭpz54K <';ro:(?N)%pFnPK4٬q \ME#QsgWj;ZhB֞F`ݘ ldoABH[ Сh^s0Ss9 =lJj6ޥgv=Zв-![.m<|tH&(bu6L7 ETe'p" Y;7.x9S?'Kx w]EеԚQ(8EG:k[.,#8\VZ0QWmgؚ\bqnZ7`1Jz2mR$)&^!T=4$;?NW }S.1 MԳ<}KF- ީDؖ$Nngآ6)":Z%pƳ"/ FWf I‚*ft " ӍO%3䢒8VVn]L"QMCU-ɝv7Iߴ3DZ.c1Z6p#HEZ>T+dD6ZZO M}'Jo*( CFN=GVU`Ike)HQp$ B[ìqTgjިc_9S:5>#+jU0Ry֌83J w31m^SQ@DNz&|dF0Ħv}#М1ƕn8j97 4"VӒԳh{Z<+ ڢu'3.i گV'U锚/Y٠G=*N#CZ-ŏaiKS-&GHB lz C$W{\K).^eESCS:VCo#8q:>9'e´!GԳM浥a$ͩVeʍ14O/:@@ MU$z +Ějv[6xv>ڰ<؉J4}3><^͘ZT \fSz{ԇʙ6r7(טJ McK-}kO䏝ޣޙ{pds)(O/ оb 7Ņ]{;C.ֵwmuF%FY G,#vVk'&| dZ EƜ9n1]S.x0֢&u'aU4wE~`LEYiWd+ !nXa]xm|Y6U7NTmp~A}| h΢x$մnP)U=M8ܾ{Gbݦ:bڵ :YhСm:{[ )١=_8#V5ꦇJȡ1 VW@՗`H17$T0.k%s¼osEeQX7/EJ+z⃊U74=,C煏묐i;˻f6ҼG?6Neܞ9Φ D-Kv]//ᨄi .Zl&R\]ghmR XBb"gb\4L :zI*G@%'̣'p5]LL4Y$蘮pK߰Y G{&;#e-#ӻ7h cv;%p4aTPw^qۻOinqLTsJ/**|@VN^tGSMVl VT2y~-+I Ł2H %\ DqaKJ[:tv0Aۆkz [senԦsBY6d@A LZHژ18"^]&SMZϭV:6-ģTh#"dkiDRtWaƺ<07]w4;t}yrB}cuE" .VF3)2IsDHu`bqRWnu1~uU<ϱ$8 i:RIYPѹxK8ljŏkZO>6;ѳt!0ʭ}33ҳ2<٤_vo~ynJz^޵ 9*\+Aҡtz?4}+S48 C#•2bfﻆ Y(-!Hi+nְ2HtlbxEpI'*kNec>'e9 o;jМ3ƌ AJFz3g9ѝߗth=JHֆҤ828]G攬K,'y5̄L3=[HC5;WAHr2817$ ka\7ҧnЁ c%du#[WA Q/*%}̬'vx#Q=-1̢au%N󖈩ǺĐ5 dLz>kPAX!D[e/yA1S@]unexl#M JǡNִۜvrG9fi 3SZ4[dA= ϧգ'uˈPx&g+7vlh<Բo>TȂ²*ʄ΍ٳ׳A!ᱷ7՜?'<$%~ |7#Lll'iz;WD4ϚAci @(Ȟ'\ވEx$dLlHZ 7J'JӵG*Cd"XgA:a']mȭuŁ(@p(qiL߶PaE[,gX̡I2ꏴD '8_z%h nҘ-SO5Xh`H\|,vO#lChWBrH[mU{B1P1o/%!CāVu `w1|}tQ2Yqw'<'Z@ya9+{k3KQ|QGcgy@qS[425yطM⿋5d$Uy1ky[hjֺ>3 {V*' XoM+י ĨO5MטG|2I|yjd's~|՝PY^eOj% u 4 O71waӜv\|GU5u3&hg2V=8/Q*LÝڡhAcֈwSm+b1{ISc QEAv_Doh F$DZV/e+V 8k Xߢy[ gX.ϊX%vOL5-}qsRAot Ɇ ܇KQrRFa`3LQS̶HX,Oy_1X['QK=q;CUзHgiʹ_{޴kִo='  n|94BFg =l Z)Y(* y,mh/hXFa*F,àc۶~wO oLwX*طev(M#ˆmVw%v<[CсZ޷ $>mEZmB,/KwU}¾&# uNbUuȟ,gl}dKD+S` xXez7'CeXZ/xL~4yv"fKkҕ'`hl. "r:}5y̢vשakf{vIhu]4΋`='|5]E9AhOh46<,~I)4]3{ykܰ`& m]hE CK2j">W)wgR˾;n ˅Y~iOmBk&檍$ .AC 5si}ֹ6]kDI>P8 fUY1Ye$Kb,F0U~'uW|Wye7rooNzJ=|KM tJиt[mPwky9*iJb/*_3NfhpG*-)[ Bw [úǍ>7rY@\Ɖ3E /Fk؎hb+(2H?0GDQ7`FeX4胔 fV%m BFgF6K BhmGQ^ ;_rh>Fd!9ϪbHxLJZHi/gRf[2P1Zh.o@Q5LgVhDs;g7-01\ƨq;P{N}߯nAu|12?7h,2vcIAh 6 MM5u-U\7S7Aiʝ8Re4;H ƁEP/㣊pX(wLOnp>ܟ% t" kTq΅@!egftDN8,,+JT/cbR?̅St3k_:qfS꽖eO! HPtXk<MGKY &K<ŭLҶjd,E8":Ma dwggT:Vt>+ٿָƱ?6ڼY.j4곧#ı24=T)<}N/TOzo93fnnQFGXbڵS'#GY pea$355g0AIj!]e竬"Ma) q˟w®B2n`B|pgw{8UUhgS: t)-F,]=_nŭY)Qx$oe&ݛAđL6N-FCqK(6z ;&[pvG `kPR.c2z4d=Q}YSGҶU m?ߍډQ$w}z~yir+9&x*@+=kgUrv ,gKO H_`II%OJU 0| ,-U9:s`GHbgTҘ**J^@>D-7C}'kWUp'(=fs{kEз(u+ĩ#åxW2S=HOVP;WuіJNS;5O:;&OcՄCs %bF,mV-{#HSCB(^^ۺ"s|͢# ]4$|xf}V&dҁ;{WdL&3P)NR1D-m(z68wp:VCQG?#ʨ%&A99wBma>mary }!v~mm-b_ +mb܍f3V <D O#RUjMݾcpwO N s"gz3JBiFoEC).!ي^;38$Od;5bұ'MLjc&ִc2V1L\'gC5A Kgh=Ag`euj3@m:B}-Z(myHrY0y9~H sp!o1eh}9Ы hZ[vF=ݏX2Ck:FthmE0* ] eYzGCms'N95pïC3LJ_HW`#P ;(Ke |N&?7njC_k`!Blg'"泎D(@@Ud&J2hkXw}1{װBsl:4u[QK\u{`/JƇ}9B3Q{OJVš~bG% sƒ-*)9`ڛ$dIs_]n6 e>#gq|w"]er͛Cz̚74uỷhfE.wp%c}jp8)"m3n?vЇ>*Wաo4*8:?9UE?P!/yhpHp* ci!^8y7*ΧwΣ%z#ݛ,buyP0h#SR_J5zH4Lj:4ol1yӂ[BoLz[o_1"8T\bСv^vHvoX;NRX.;O80˃ 3):g:Uo!ll}ycCm{9)d zQ:2 U:e90$Nn /vw,}ۉ]81|AS6Λu 6W5{EG8DU뺧|)D)a o1֫is>,ԵV|fYw#tJ`jU9inX bSZ./i-r \v!Jn+H_9^ѣo=Kl\0AЦt; llX~˅+ юƬ :_ f#\V mI"9pGP CwC>-檵C"zFm'J&X+3sAY6r; 늌jX3z/+t$f6e|0{ uha|y$Cq!;NG܇bf6t)N}!С0!r_qD$ REk@cf;s8DB&Ps+Lvԁ4}S[HrѬ{H~Y:*\^+J[ʸ -h@=CYsi-2 Y^QUlJdSxSJx.4ȟq(Qr4} EJSnؒ*:B\؄qjq@ Uo^tB"M ["_@ ڴeq2F=W@,vM]ӡ]9{EJ UI3DpڄŠn 1\Wk]z}pʙ+kIRjdt\B۫1vlh8%4T5I=kC9Xj4kOb=swh} c)PTh:s'[H.|. d%n`˺Cpw7xJIbfޥ'An62=܅Ż(n2UQeW40P 8^vg 7P8ߦ]z>^6c3 L<3ZX־3n撪LqSP5&Onkp'8W``]f_gAEC FTԍUkV涸Unt_"65>Մ\D nj-0M̋XGtC/my ,FT)0>u q#&NxӁxD ]/^ڞQ EiߑI=?]\/ [X/Z .rkqgE^tJj:H@p&!C(5ˇ_>KĠY7Rjw b mɵU'EP &BcJB><-\Jm73%|;G̪`YwAEG҄[׍s%n%lB2q+;Eg-j\aWղ$НLF>D@G*Ag0"\ Lx|SGu h4&BfLRpBm䕖hPf5Q _VqO0?9\!=ʈSX7!K Mn\FPe%LTysVK հ&;Xke-ɾ㯤cbx1܋qeQx7`Dй-j0vuƒ7c߄M I#ɞaenU`9Ign&`n4o>P,XL?]c-7IAOEe&`w"2%(Suf\&g&)ځFY &!Mi^e{VGP(HdN_ysca7g-~G@CG0)ęJWCEJ[M@U_D9@. ".ÍMOH0s_=P%R!z.:><. 7*J^fc'&sIdK 7=XP*:EA(WZ2T!.u[/ Ew ^#B7ҟ ]Aw}5VXKTMۦMA:y=v@|03:%\@n4.O̬Vι}c?Ȕ/4c:O@uc ȹb>'v&0Z (H!/ +PҼoft={흲sN""1Ⱬ+#8յ#ɜf*Nx2ּ%iki<8}5=az0z 6òI5$z)E8 IMYum1y:mrXpJoW@yV̑{&ieJf<7q5afhZZa ։,>HEzw~QD kDCu#n1ιE@9Q^r.HiTn%p{=-eR7"NhcNj774+(:ҵR3RYT(!Z Јz#f4c6QPV8ɤ]wXc-B;VIʄ'ˠ@v`.5Ɲ;Ļ5津$<^ܠ7A 8\921 ~d ܸkTE@Y&KO-Tqo߄/A,Ѹ1|0>txoL/w2blXa!.WJ2zl&{i)؝,wQWQY֦by p-^QЇ)9/l.pulh3l99M *z2M/}}@rs)ea&6ZebrnBgFEЃUkݡv)epv (T]awɘ`~﫜rNyybnՃ)bD`2-Y Z>oBP7!s݇!"" uCȗϢ&[4  a4Mk}y#b8E,-ivahN40[@z 썖B;>+{}q+}sYSiֻԺVޛl7NMRozoDLݡɫ_ң71&iwERHPJs7x$o `˩k)Hf7suD1kƣq#=MyћZr0|G.)sm[)+@׽{`j5VAd/mU~F7OtP}wŮVɯyϮ}ET3B`X A(E @:#ۑL?&] Hԧ1{~":j4\,gs.?:wrM8ao H k{( K6-O+jp(>-65V>u:3ϲ}ߎaM GN#}oQ̏H!S»ۄF֩>N9鋶Fzѳj(eN$[t qk;AXU^Me-p`s,tnv" uw_$A@U~/ctM Kf䈝n±*PUo]BoG UU ޵+خbʣFdD[{4˻b#2r{M/iA&WdrEf}ÑEgMA5M;n*N.HQvqmq.)㥤#Ywc ȑsyP dQ*bSy;txS$H6{ıDRq}GPWQH:=ŷ̶fj+`p;)ym4 3[ԅG 4$b Ę =yfdc!wW~,_!7L]3]SdE3dbOA?El+9u:Dc1z0sv0̶[BR1~{wf$20`̨֢xuD(^ 01s`6kFr)F HeJ;]]Qm3TB>vK㱧+Mpʄյ ]  T48/%1||&(KHɦbHrK? Y84eW(j!R^擣otzpG ȕ,MDq`LK3|Ba6eoj7}ř0EN( tHe3~AR6HK?*|< d67pLq'tu`fgУߟRf_py+_i0igcK45#**oCx@ JVv-w=TNqnEnj \ܗ4>e;جjBZVl iR+N~쓵0@Hc{4sghr$#L"o[SFB̳Ŵj =DYsdPݶbW !zVSV\cRQqZ?厸[+q>֩¾i?b|FwX dnZk,L= 15qЈ-÷hgH^t 2׆ttnSNⓞj.H2Й`! 'b&?mUb2wY ʽ򍅫 8Q7Յ59dgvwBh+ޡ "Ի@ `"dcԨ_B.։̵ˍwT8NѾ!!'C ^˼(tYM7'ᛰׅeMN 1Y/ Wk-a3DR:"CZ&k(\K=GDtN]Ӂ)rIȂ|lI-'bUY/D>p,8.r>.9HmT6 6ao1 dz;Q`Հw(7 @i/+Ow y"H,c1V'>Ukܢ5ډ̦ukƆ!'^dV#ƪ?2QvZňz T9Gu1T"I\gj;> 8Awۊ7Bg\)T 3q B@PX@cuHL\6-i%٢]VAd ݣgI;C^,m [3^yI6Mo#9]ć7,Vjxdق0=>²*@i 9Rտi;8a9Oo^ z$Շ^0C6jr37_}Zr/>5OڱոNǘ;Ƴ݃.쳷36hԻ0n<ċfU+1G9 x_0RUxcab)6Aq4+'ҕ㘏0}zLT2c&FCɴ<g=j`pD\_*@}RJ37Ǫ҉-!bG@8+U+@Ј k8r>v14XϻЃG)"&/Ͱ&ҳғҀ+S_}_RBtA5C.wsj>hLbW3d8g0Ŧ7z>DqͤB'0?exxO6# Gn-o/m-iHHo="~^{!oOX%j}>Xt;6MH_8 !#f3G㫬=;'NnXVUDŽ `dSһFb0?AXeL j9!S:-+z g~ J?>'*eVЫ#aN7)?@݌`Rf7zZBW_)Jz L+r}W~Ï@ wtv^.J )FʬY}E^3mZ:99 `pH)nL@xubYUODKI@7}З3&W 2$c=h*U7ud،)WY44~v^}E5A2 ;_[~~Oi {pyTCJWKbz?7VE 8ECEt:Q T _Quhjڔo4I}c:yE! $WٿiڽҤ"߄>5if4 ,}C]_oW^h(ׁFTG4 gŸOB T)_o^ c1 8zKf.Ma?g:5w\;0%@m'-(^.rᠮX^wFM_&a >>`t 2#">όySj.tv%h8J=@qv^c7 uM\)5{fEJi\k5YCp%Xɕ fJʇv`` K`u<*J9$m) E҄ Ng9t?,8Prq?\CR]BڐA$㿔i2<æ,U6 kf^=ԞS=m#hP6v&E/wuWW:^_y!Rz*T- 3o> (3_+f2=ʏ81[(/Puc7TXfK71ISzx xR:TRwT1A4{tk3wT$]S} >tRNi/B A51pr ӣU8MU/_4R~zہ?p7;Z!JRL {I:P8Z_ؕnؠUr@i:~*C{ή }OPhעfP;>o9{#مzrPzAb3Whfg]e y}R˃m/Oa(QRD~A.FJ4oOsӊ62E{z4ZH[]qޗ8% rs_i),~`(Z]L'>ϣ^}ϙ+^|c g~%^#~e-M(sYn1GV vk5=@T~83D}Q`.v<̗:eit<Ýcm>{ MѼ_]?fG|Ȗ3ĿXH:x \0fZb84:4PigBIy^)fȖ\ Mz_}UGu/Z ɚ>Xq2Јw[V0o%$:t_]YB(yUGv\A01wkj5;"8 /M84TpE5bzIq;2=/|%S*}A''++}\I٨p7cU}Q}Ej?`m]RB]8+dPf kC')1Z$>+.؇]}pe{~>ֿc~m tqՋ[./Hw FL[3๮>#=G[gtnOSHփm Ï xbm#zDy|րnwnm|~rmP,ev1v0־:^0,sM/d4.]k醐?PɄ>uc:FWEV/L~}_hmcv;d$ ۝i1:pYo>,8ΰڴ=M1׺]$L<ȃm{hVnt9![{gJ %:…(g1'9Ak$W:dѪ*/>Jsm0گb6v}_KU]ڽv@;s,fuTnV~}2ejnazy18ѓU}ТWGdM/GWɟH03=t@ -sh\1)e{ytMd`u|&|d;͇h) mƛ;!.Ur/)u\]sc/IuoswK /GI&Ȓl02|cIx1@驿.xEFw6#Ϳ_omYjNX3%o1" +++kZ]W*י5fšYc:D*ṵeۮeuLn0vm>}ӁߪR_nFӀ jؤ@_ fCp?'Nu ~%{4X8wR jpSm)q9ק<c j k6ioϋU>  ,b@@܂!pG3:kV®bbDF]IC\#iXw,W-v,h,运%% {^ D-e+!Wr(z.}la>ZTl&T Ni?-i[qv')#|5NmxP@Û2–_ w;ӧݺJd/2{)]0t) 2@`4,@jS2a68H $$ȗ$4v$U=/)'P;W6D6P _AmTeMlo M'IFRD(%_!YWF)x/\v"y$ź W;Tat1}1Ƥę~qg!XM0FdW@Qg:ثplQsA-.'fCeǔ\cbO=?(g~:xye~2t& J"|ݦ9laךuN#;oB,Q-9}ˆ8,2vȕJ oE$z;gGgqh`KJ镰PM[JB m٘J[u쀠e9na[Bk>IT!6̺;asՓDtĻ e.1Cq'26nՇB嚼;9ʣWS^$"w$6+LKRI2'texr'y$*./*/u[>hbM#ƀ#Cb_MٿPze"lv3x5Wl:@6L3L8lm|N`C{RK8? tq NusbFi64)撩)uGAuYH]$P7InewA0˖dc+6\H/ݹ ǞJsr u+.CRh q1\H%bDzcL1ԢĤT[IKw)CmX>A U/| 'lC@wثR`s7 w^9mk0qJ|R`v+^mzS3 F~!F"O6tkK]O\dBh18:浆m\i욪OR|L~>W&#KHI6(0:U6#H8zHti@cu ~7@?qT{cƌg-1zxW\ΟY{lw笙zk/q:<. ɵ=ė4QΙ!PPj+Z?{G(BPRw&ÂS$|W# 2DR MM(,~8A`7Xt;V,Qr5#) raZ+%RfXV!pqSuWU&?T0>|Jɵ[_\ov0(uVk3 K/s] =CEU_S`@iWRʨ&"{D\U4h )KTDU pV8?אقsĆ:l7 E_4Ňݚ)r#d$Y mTYu?HմKb]˷j:Io[^;ԗ)K>uys$upq齋ZL>'=bP͙\i?T mu19UcAun:p%iCX #H4AՊ` Up>5e_ 7ö.[no)o,Ӱd8$ "2Ŷ8ߠ;M5e dvN9igB+|]$!Vy [_8j x *buVr7`zQ4lc*848$͏.&b"ې %Es+|U[>ٴ6zgRm׹A6E***.'񻴟<1EP!ν |ߥSOa+- 9@rS&0$/f(m2tE:˨[^TULI(%Qj6^hP:MH^rcҁ&ȅ'~(sIj#Va"j0kFk\_!؂;JD jK^+W2 *`p~IN2*B7KUxD#mn}DL~ ݌±QO(R(/f̟ĵvy Dkp͇M012rE$ VL8R} u_N28M=ajW.wQd퐯1 BlHG'm|hv2k?b^g <]7GLpL clzpy zoKSKΏ>r* x\Z!eiQ6l[1:ώY. ?d3)v$vOh8gA]7RU=D)ۅLjLrc%GnYg(mV""QMFji*-N}U;6#L!d"d|srV3leVQ!kiYm &SVʕ B)ڂ%yG6Bqח$Ksz$GM~}\Gt7Zr{!6'\a[d|ûWޣ*)"|HAfΨK>16țIEɡ꺜v8և%^/Ma 3kzNdaORL2uX(vGk`NȊ0#7r ;hT-~1!pdX pQ2If^`mP*>H]Wo#ct &S͠ȳK۠ݮk( _vNza7b\DzzԐѣz @qB\@l5:&e9vM6>NQ~?Fы!@8fpA)E~㍜WRFi1XOCÈhި%92IA2?xRxX2#=bWV'* {d4ySoŚI U vOMLwj*'B;JKfU'PDf L?4&W*e훌*=.j'^ Us6iitC@CZCd,r9nO0 j M$XM@ @n7&,12x )Z6/%L #C`mLf`i dnʻgqJ@ꓣˡ) afwJ'`a)SYÏƙΔ0Nچ F/9uVv,dL0)*#7Oh?sv]D ~7Z2Y amϝ2;(?eږ/+^7.;'<˗^Ms0n61h~0/ YYB4`+qP,?NKԜI_ŊX2$#v#[4D*=ip:E~TvI.yg[~)~GO Cx+c&=8>rY-6k{/A] jAw(\ 2wPcbQ7tх#"K#2EJ,C){j$IC*xăW.7CI~D jF=Rڪ.ؖ#&Ny]0-RN lOlƫas'7~[A~P/:=͉=X᮸ڿB9n H3^ ٰ}FJvl"Ik"YKo4l)X(˕/wQ=_0L7. S_B-_aJ1$oj׍ʌzo f,pt'1xka]+xF.eoV-1~l?/eBU7kh3 U*aj)/fĴY!Y6] *qmV~ qaT 10߰{mbB+AB*!ZU9XR""RB 7] ~gt1fKyVC?a9ih_#XN:٢Q_qlf_LrGCڻr?vV+W$r0l88'l͟q:.6=L7@<$ q 4}~㺿!MMB -KK/+vLX`Fe>OEDPfЪoVR(_a;P0/!Ē ZT+QnѦ:>gILOX䫯2 8YRfjM@r5X. RNd27&ГS'abodtU7\,FQE݄h]2L@)/d1 c??)'ՈW<)zlGWc݅7 ekA6J(y!Hq@҄7m> 8[8Mw*bm|asmƆbic߀@F %Lğ7\p;A0n9 IG6# Φ^q9_1 @cxK%H-cf$)'>(5|ĬRMQz8Df!%KB:OFY5#kL g 뉋<)Mgx`)pÈ3 XwЦ"(O̪!v "+9}?~h??@ мvHS%s :#y9f zG=67D@;;L诀2ܛƺn"Q9"z F`~-=UoƷQo6#)4Wu}F<0>$Jo3-j<@sl2s` N'3ӆlz-& ,{ ;9VRO0yUdu_sh`nShD* -~,kNxx.dm/ EA1X]Ik%ہ1 \Ly5fM}a_>肁Y2q0o_mPS۟9tCsJŚBY{ V?nAOT +&Uk7T\vbVI`ے~eLa;[:ٰ4sf ye-w R }]|A5%Oy]~]F"d>pxŗ{xd5[ʹp Fb?Z2(YlD qMO"!zZMJni fϚ_]A51EMC ձ -i_x@ruO9vLclYS{za},xíkZr$ 5g4w ֍0|cỉ;f}-Nar̔Go%eܯa((g3@~H)H' =PA< 2EJ%5=g"Yd` xgXwW؟  ;Tw.·(?ǦbW0PBsT7QUA`-w{CW^`@`(dAI*va_c ɔ޷˘N?ǃ@ogZaLW[ 1<)ԝ*\q =@j*+f,F~`f|8(Kax nmR[3Q,ҦY g#!vۏ>F>GYaKשHz )W__)po_4bt{p\ni4vֹS*3(gp^ >5كva]عǷrmu3JeJuɰty=H5=׆d؏ ~yЦ$_YNq*,/n3">;밖⃹ ps/)vYbZ`ґ'K|\CZCDw#s鱷(?61ՠv,|q(x23/)u cNw92PM^3EP=oyt c ^Ys"}* YGh"-ɱz=$oBs+j&HHvxA5s-F#jI|@``x*!~.y"a y{q\i/DC띇1w\ rA,^ 1r\]qk6.-Ei9}F#T|٪9h<0A%ƕBL3P/"+'p2N3|ŦWg^AHkY nklO&psN4ED@ըʊA&LXc9B``\Hy{ϟW^䘭6q=`Y.(}[_-Y7 M:vk8 +p[/53W6Qg 4i`@fx:P 6)?z1Ry~or6j@tJ:Єb|9cX.x0,/:Z8zû]V\9Qi1"?]E&U*p1kG\am9\ka8Z=*̆{XW10ƴE!}cn#wUu 6S{ݾo0"oML6<5\ZS-e E]{k^*!ւAZ/-x~̘y+_&Va/[C\`ڠ:6Aw`㲣ͧ ψGc'ʹwkM,CvV{̹,8.:mݺ.l6ťN/˃g NX NoHKNwj9yʙr./x BL @_|]\u[h[@#z`1g=B?1$Dy2:wA.JRNi!㖝Aj{ZJ^K6v4uZa/UIX()ЦͥQ S.sm숔a? <(V][= MuV QV[D/ @w_a)p`_XuY7ڞ]y(/AY>FNvt枅*Lp L9#=65l': eQ3ʦ.XQ8emLbdG5|<1B3:=X\^К <EKHbd!/#xb6e*g7PUMqr7vCeFI0$65:`w}FJY*c^ezv!41̈́"jp[ބ?dtnR|uh zN#`lX)1X[81|x1_X(o=sJ6hՒ3t'@_i s70ݜtJr4]A>n"TQt5V#n `N2MHZ:d]ehޔѶ6qCA6g9UAfj#tVIAr G/nV;WZzo8XQ g0}2h?M…K=\=82ij»#/ʽ*ݥt<&~qS- L;c? FI"&.|w.=NO!\0KvMd|‘v.pBc?@4T>{a3 r}I8ˍu_ɯev8dq:I s.u=xd!g.>G|;j ٦n4v Ri''=ߵoͮ|J !՞iŝehѶ^=췍6)o cɟa؆49׃ ]g([lx5P@sĢ׃_3qxh<@yR>6oqI#=0,4 ^z+~tZDr>{7|2P 5u rLA:9rfO T`{s= m:K1\O+?6 X6* YeëTBۆKWt?Do ((oF̈́.6!v9m,٫wl[TQ(fQׄbhǔ -`L0& &y3{<ƶaiUE79HSmRGˎc?>0ܗ(~ ތoQϨt^< {ĆFÒ9TM|l/m ]zp1@©~|(Xa`Ðz5cR:sʘi,N0R?p- wDmh68j6V9S|cp;O ~S tq/soX~T 5'|/aK]!Jؐ:trn)hkqjY">M=x^qJ0!gJVհÛA @&6;N|Yshr&2HrpSpZ~b|0\Td\6aβpv`eX|L%N~*Yp|]Xݢm%Lf:!Lq5 }e+фޙsjl;!q'? :[TbS8lul\h޲pۘ=P#N\(wz[Yݺ~JT]J1L.ʥtښ 2*lHQbxuvy[3̣V[zf ,#t!(֗c( yF,#ӏjB;˶=]Ӝ'-5hiUƟ7/*ʳè-ObYry@6niB5NʩcCꆝYΜ n8c]1=cb9mkB݃?s,8劧tY⮠G*IDc3-Kg{=Ed\@8]焇©z6nS /VbS yZ>C0.o-["ێa1B$-hǔ슿K 1[!l/c-v͠~ʣA\=ec{ %k-YFMy{EAJ{`8j{}YDৗ0jEř*µvyPfm$\RM3Tw,`("ɉd/?ݾl&ƴlz? . 7Qѿ)T.nl3[ 2æ _@7CvJRxPhMmx1ciuW *KF5%`=+c6Ss}O=3" [)n%vͭ`x?xlqM/G{e]~1nNAb` 5z@@W13,%҆74|)֯lv| oö6 cR1ֶQAh)uBX!5&Q \*z6yRe"HJaĵmPA>WەaiFOͱ!۫46IB.#.q쎲wط_ o44 ^a hnu#YP-F!L×Ì-6n[[l::!\~3-n- J?P žz)+&ۗ)WkF_;iF`1}!~ q6?n&[͹hl LlغJ: A0ݹ9qGmMI4MyPyu  eeŐi'_ZAq{Z~l19 \\ʚC(B~0z'z /j.$O:q'?,@Z*2z(mSI⠂y<-QP}Ѐ}^F+)l+Go 6 ;?J%.a?r!f=1vpΧ3,ٻ;.ir|3e?hSNؼeѷ6uk"b)Wbou|@G̫;r;{IVA:LCtXy ?~? Qsf(W_o{yt-Le.isP JWc0d#5[-{^䄪wk$ YQ}C:{bb;uoNRPzJ@n5 ;'.[1$ u̟Yoj-7ƍP!+l>+mi'ȃׂM2  tDkJ9-1)Rn\w[4.] vs<5"J%P^jyA;2B/\* :`g~QV)ʟU/xS^пk2ĦWTb͝w'aqQun?\Kae3+O9[J&Q2;Y)9 y2hX8 XO) B\TG<_ʽQg3Juy;b߽B۴Ԍ0}^yos!kG:p,-<y(#uHF67q:avWOZK ; Aro璮< /ƒvuI%xzYFBn>FZVYH JK$d )kh]6R0ivecX*mJe4' o_UEz! ! ?ۚeO^?T.c."HhlB+2قuh)@/ c5}?-pxW7^ٞfѢnjyi|&MVU:_QYoVJ+0Ɠƙ켡j*E9юZuV G? f$0Cf#꺪WMQM9{؁O:IK@`\1ҜܫFp!U!ЗWni詏r,)- }HVovEKN!Y#o"v5m+ƒ˒B:ŵ<[R9Y M@K/NlK4Ba y7C 'wiM`8D2LAIFB.(APɧvATWd%Acܯ"ms;R<#r*~AMei,N8\ rt.jƫ>O#X(7#&@8?Z䈟us/k,sb2c]2OY{QfuOKh#~zx鄆 M\]g]|y[6nzL/#0j\G%f{$ceS:AP"%~5EvF=/ἎI]vaVV=,DBN a?j_\Ǖ4޵܄_fg4n.-#vCX)xq.(T,Z2>0&L<J&CaB$V_S+lأg2~؅YmP,QkP̭}E.ˆyؽkly9MaWp-uq>^&f5 .j@( * >)"aELorcZ =X46Mhは.sAT,8(:"PԎk伍K~`*)=ǩ$Msmw3.⁖y,c}gW&^QOZ?J*cQ걠[i+˶٩n8ZRĂW 6Gߩa1q( & 6d' Z2`kbi&6B`NV}!,m]H2uݩ$oK EWeU Cl&Q@G4^YCdcּ,kM&e-0u qI[NebYllh5fUi4YZCp]QArHQ6ƌ/FE;ˢ@cǞ"EsŽ)%ؚؓ/_ZcwX\C3D.]nP6Z}uY ^8<֣7hi ͈?th!4F \~M7ov@cN }MH*rk FYH`<>zLNc?5:Z.x * BFp1迈>h9'$*59*Bo\ Anݪo22fc3‚Ln Aso K60LNz5tPG3Ɣn(kxGhGv/#1 Zb}6ND} 3Җ.([BZ|k%!t9%-=SsjŴ>vY%k $j,pcő׃9A߮>0lzkK4TF9>(f&e ;QpNF޽##@y'!t\g.YBHDu 2BJvjr&<`o;x *$b(E*ud פm Kn4J22rtU4~cHoi# NE|_lFKn[:XLL>P'v4ٶG?\Sp؉cޚֲ ea҃<݃'NGsf ͋8&y5|k`ȇj?WGمXp j$^FvE8)/:zl :X\i;]ycJpĴp-]hy30f.,g % _wx4's,u} WL#A*gJLʼ8>]#/+&ab^'q4hk"/4MzyVrY싔9A|pvEzITkitc3" ]5FiA4aN]ħ%w͎k¦Dbκc@#X(֟:h.x!9f5+A)/C"Y:L_h;vۇN 4 5K[$޸eZt Ae40BnQO[m8dvt'`,wDLxpa+s(AЩʄ8/g:@L[yL+F|h:v:ӡ VߐCj94Os_:勳Y`ONW~p|ܭ; Q~k{4)|^<svHbh t݈=Ě.R!: @%~  ᑪ `ꏃZ&KTe0hT* QAdǗWu|edq6fbt7^`XMnpYKVk.|?^kc8 v (bF~Y"㸲 }M;rVXVJZy  ؽA摺 w%y`1X{Q Z8UKi>K>ճrZ51ݠƲґk%"B?>Uʰ*/.28,tT|#gHu_?(sҁT{G†s%[(tڄC03yc!1wGy7IQz@Uڥ N]"po0`4K΀0=[0ވhaqZ<su @԰Iv>ڬvm6Q(0¬l3$ *Ψ؆H ~DN5Ä]J[W|Gw3-A"-2fh[Dg>̊6<%Gׯ5d^IM~e? ؠ#XBqj媕~dG?t{˱KWItdOmo>Τh;#wrnv+P$; {r&B|45£u{.d/*ok'Zp)'hPX׀`fKGk洁c%jxmR[ȍ27$~!3n3STcM`՝42YZ0o` 1%(*6RD`Hf^CTS0eh\Vİh\h2EU&3A/fyyc^?[d!Iv lDHrn&f*ًa&h#`teRct(8H#WU=L$ OqPڸ~1|ۮUf=ػհ+tVj4iWw5֍ۋc|@n~5J sh4:p\~9f.]|D<@-*Ch ,G؎(30ey'fBK'#dӜd1gPxu&]CMazPVGO﫳{yE_YZQkKo ӓjVMoh&ng?umiA=FOQ'K}6<ЎHߙ}<wEC䣿SG}?,6rڍ8LkcP&mrDV 2mXy$=k'ɸpR)Vu(V<0n!/EeCzL7S>}㑭^?a'}[ ƱSH+*~L`#Oxݵ;6OIjϷ;!>NMؗ*2@,3wZYL6LMC709q $w[>?>¤2ڦڻp<{g=;t1`QUɨJ&#ֳjH5,x: } o~|h~NH2u4l[O8}+1< z(M#ic:–0p.>VC]P!lPrebi v/h&L=7셗́D3M*dߓ">y1WkݘkwwL;;I4qm|^сɭmf}j+\} 6diAMhMcAIl*W4'$idMżZ0e;3%:2QI@UWҎefa@OeAϙ; <; MoF 9`l3W6_FvJT8K NaxNdH bnPdE\GPÅg\ѝАqpDkTok2a_eΐ膎g) ޿-z!A`gƴC<`Ue&yҞmHo'$*7j%A=bac ?U9qkTu"9)J?݉Q5O3Ф̧>N4<~G nb`zҘ)z4RЮ$7@hP٬WaƎp%+!تp_?ڐ"ؕ92e05%u8U!ۧ;QiI|󄥶<z^u[,y֨-lh?u1YRhh^֎B [iȅS7IfN[N ZAo=;*l$CTh .Z>j%-Z% \䒐DTTx1%p1N}&Ŝۇ~TLR q %CJڧZO(Yn@@ҭ9IpPps39Vwu0"F 9~ǖ)^SrjV|ʼc{ 3z'i5l./ 8Y`&"o!L=W=-9QQ'cz!7//_뷯m~|^+?_??zgOihY\X%=§Y:¿)++ѫ.&LѯGԦ<6cN\f uF~/.,Ƴgf=3޲oSfĉMơǬe)Z軺RڂE!2 M@6(RΡ d=H5qه'syRt׆[C#cs߾OOU@o'4I'K[w6ՓhZFc=1?-p3#,wxgmL1>-eN'#M=cho?N>36i+94٘ v.yƘsގ&K-XOĈJ8AҶU@ECRѱ;~r;r#oĈpZ us6 R2%F6Z: ,Z5ۈJ}+* &&-QvsQAEM(A[RJ@&Le^v/̚- -|iV=uIؖ/K\7].;yee"ǖgU&^âú>Y T_b/,e_WfD$`~oÇ((FLxZSzU]F^/4'׻ƹ2CdxsGW8K) ]ߊ!Nި\{^xvV:iUR 3"ՃKԨwN^h[X( -dMoq %utC&iڦ`16FԌ#S6`әB m9XNpR\ Kt.n32@VOKg -ۙb$b Z>ܞjʎഥ|۠7ӧ"".w 4ǵ*1L(Yl J_1F|sy[U3*{#q vy'zy?:dni۰;BmAo]nM1j@1'Dm/GjW',G_5BfFQBzE(Q:>7lE((&^k/8>];`8Y iPUX:g<|;]jڽu7.bv>_yIh$@hǢ&-; ؝t5wm;(܋;:ț;N?܃Jǡ__Fn`mUʡAeᨘofâzBCbV2|_( ak%˖-ݎ M80Wx,J A ofw2kjp6&9 !#_8Ls䔪Үs6[iͲl.ziS\xN 3VzGSVKe͕XX0uP[_Fk;%۱,R[tx4Bal!آEB.OY؝1k\IsbYCn3p? ίt5.t;.>;h~Ѥre @|,/6ﲶʛC?0fo=Q!' 'QV=#{.>]>]"ٟ$$jkAw.-VXgXP֩s= XÁ\C6Xĸ)%YLOD.G|a۞`n4VP[.ݪM {v( __(2ںF'={\ٷ }2$%vDdg&MD4.2c$. /J Zj ߅3/ռp€v0ݠ'pn>OZ20$~l0{P)膜{l#0Ɋ/U)apм;W㵽%Kg@ eZnZSj4RQo=Ld5|H4NDr[[TމtCEJ7tҎJD3E>ܼiܴwȃ\5;i!ӫ3'}9:Xu8EҰ;>سٳ<)-~JA%f? uC{\.L\bmr:O{~vrr˵M[iBJ4AFmERf8fy6rY 8Km=lhb[ -'׍<]珷nRxrId\trݧ`dޯ{Iऺ{&սyE[0rՖj]0i@' Ǘ`f22GXńݨ(MSH ͏3ӜTwj>Iu?S2a5{8`~Dp bDx9=Wi_AnWm.tLJo ȸ֬J&|dɒO7-tmlMVF|2fG*-S/B +3Mjz:'m+aHVkq1x6B|g>\,G_IeCF0e!$t;$ i ډrhfyo,"&HJpq;2]==^~5Fy,x"Vr0I$h>> wC g֑[8O|R[1a,ހ :jbMՍ^]f3SآGv~}/% C|B Ò 1=)> И? i` ϵCRdY~"}Yz[Xja?Zط8\+=^oÓkZ^Mtu(OIO \zcMmg[ [H0ɨb|ʦG~'BOr&-<ڿ L+BΈ*\M*%mZK.Ʋ@oMNS XWJ <E$(/%a4FD5<۪ h!ƂTZX tFP˖]~͜ss˯>&]Woevm7?2P$._uw]?)*)c-),^/Ae?-GوFmop)#n do0G2!1,.:NPn]W?PPO^!87/u5Ipf/u})ϓ3O!@sVX+y#ݥ3Nk|6ᖊ fSSy t{/55l"5cBT͢c"mn{4r ?$ͧC\ )]c׆ O&t̅Kho;/yJnN58dҸ5|IN\'Ki>xSi\'䚳as2g]ѫoR}0ݍd^8!Z¼wZSckt3ô/WkgݺF~sn*v#!@kⷼfzv_Hn:xYƽ㖯MkJ,wb +ѡ[Wa5O4dg=fG^u.0dCnE]rWEHiy\?3#'67ħoZt+_3/[$VL  _l4(Z>f5=d-L07ܛ#iZO&N߆᜴96Rһ8o:Y':AvIXY}_ F" 8ܽsJs2רW^iWvߛFG' bc17(w U $Եu\IWx\DSd dg4s}兜elC_^'Mxtջ6g WgnuW9$ X675~iPy&GHf9 OCnӎfv93‹u7 u*D=Uz%E632d8ba*,IgRм?CNC\+Yd'MmBvU7P87n0K6/хM]:7iu։\7} ;'WrY'L:=QZfǢAaw_<BAGG?:Q! o+`Un_Η(p#*[ܓYL6ݙ+*GCH ?82]w4Mw,= iݣ &ALXv?h BZd$o , &~* y֒p4HNIJg `͌Ri-ܑ.tnn{]>fLt|t?P`">JE>v֮ b0 @0 SAJvd.!KN#ⱂ 0E O!$-x">b /ðDOjAT|~kv0Kjyt6[܏RWpނ0mWUYt"%A8bHBQz 4n޻j5IENDB``!G,x]mEOSs-' jxXlW=믶0t" q0(`)(آ L@, iE: 24!CB usl.N30F\Eq(y F^9;߽{de3 Ҭ}Ȧ숥xu]vmgF6d݈,4DRB?u`eDb8}8wqNqN8p~qxgld_Q(u((5Ŏ,a&2lmь,DG6܀Z2C)ă0 (<X&tspxޑcRҽ|U|߳O4\kM˵3zb^ z^ )=Q줞pces̈\qsvD+֊y-0g^ ZQysC+kx-0g#5!^ YvsVkGy-0gaWZ[2D-}Qڑ^ P欏vsU;k9vsnZKZkIko#7W}:E^ N6BFk9[S-ӹZ`JuutʽX} ۙ^ U欷sMUk9s"^ Y'UOsk9K!ԏOͳH4"-ԭ|- opG?~-aE 1P,>=66#l>FBmk+Uȷ嘢X`1֠P;ȪnOfv̴gFn_`\ۋEK^Br {0kXack*aU8&v?wc'j=lcFAC[u3~KRX%/r9ﯹr!~χG9lYg(g;nNKo ^R4LLx4CPgT!eZS=Y羏-SX=OvT,W܅ݠzy^gj=' ViBP\m,ig8~q,(FD_&3mjjxU+]hx}F1T/T{_Y6F=XC?T2~iD:eWKkT1~iD&eהkJkV9~iDe[גkIkS-~miD.ܥwR6/qeesd_N_b_n/7ѯPBF1ٯ0WWl W+IUQLM&Vv7;Xobm%Cµ>:Ǿٱw𱫕=5o33Gʨm<>lkˈJmMn}DyK _Toc151540212}DyK _Toc151540213}DyK _Toc151540218}DyK _Toc151540219}DyK _Toc151540220}DyK _Toc151540221}DyK _Toc151540222}DyK _Toc151540223}DyK _Toc151540224}DyK _Toc151540225}DyK _Toc151540226}DyK _Toc151540227}DyK _Toc151540228}DyK _Toc151540229}DyK _Toc151540230}DyK _Toc151540231}DyK _Toc151540232}DyK _Toc151540233}DyK _Toc151540234}DyK _Toc151540235}DyK _Toc151540236}DyK _Toc151540237}DyK _Toc153533899}DyK _Toc153533900}DyK _Toc153533901}DyK _Toc153533903}DyK _Toc153533904}DyK _Toc153533905}DyK _Toc153533906}DyK _Toc153533907}DyK _Toc153533908}DyK _Toc153533909}DyK _Toc153533910}DyK _Toc153533911}DyK _Toc153533912}DyK _Toc153533913}DyK _Toc153533914}DyK _Toc153533915$$If!vh5}5g5d#v}#vg#vd:V H!,55q5u9/ /  2 24 p$$$If!vh5}555555#v}#v#v#v:V H!,,55599/  2 24 gpF$$If!v h5}5f5h5f5g5g5e5e5 g5 e5 g5 e5 g#v}#vf#vh#vf#vg#ve#v g#v e#v g#v e#v g:V H!,55h5i5h5i5 h5 n9/ / 2 2 4 pւkd$$If" n< p:m9!%hihhihhhhhhnH!44442 2 4apւ$$If!v h5}5f5h5f5g5g5e5e5 g5 e5 g5 e5 g#v}#vf#vh#vf#vg#ve#v g#v e#v g#v e#v g:V H!,, 55h5i5h5i5 h5 n99 / / 2 2 4 pւkd%$$If" n< p:m9!%hihhihhhhh%h%nH!44442 2 4apւ$$If!v h5}5f5h5f5g5g5e5e5 g5 e5 g5 e5 g#v}#vf#vh#vf#vg#ve#v g#v e#v g#v e#v g:V H!,, 55h5i5h5i5 h5 n9 / 2 2 4 pւkd$$If" n< p:m9!%$h$i$h$h$i$h$h$h$h$h$h$nH!44442 2 4apւ$$If!v h5}5f5h5f5g5g5e5e5 g5 e5 g5 e5 g#v}#vf#vh#vf#vg#ve#v g#v e#v g#v e#v g:V H!,, 55h5i5h5i5 h5 n9 / 2 2 4 pւkdk"$$If" n< p:m9!%$h$i$h$h$i$h$h$h$h$h$h$nH!44442 2 4apւ$$If!v h5}5f5h5f5g5g5e5e5 g5 e5 g5 e5 g#v}#vf#vh#vf#vg#ve#v g#v e#v g#v e#v g:V H!,, 55h5i5h5i5 h5 n9 / 2 2 4 pւkd'$$If" n< p:m9!%$h$i$h$h$i$h$h$h$h$h$h$nH!44442 2 4apւ$$If!v h5}5f5h5f5g5g5e5e5 g5 e5 g5 e5 g#v}#vf#vh#vf#vg#ve#v g#v e#v g#v e#v g:V H!,, 55h5i5h5i5 h5 n9 / 2 2 4 pւkd+$$If" n< p:m9!%$h$i$h$h$i$h$h$h$h$h$h$nH!44442 2 4apւ$$If!v h5}5f5h5f5g5g5e5e5 g5 e5 g5 e5 g#v}#vf#vh#vf#vg#ve#v g#v e#v g#v e#v g:V H!,, 55h5i5h5i5 h5 n9 / 2 2 4 pւkd60$$If" n< p:m9!%$h$i$h$h$i$h$h$h$h$h$h$nH!44442 2 4apւ$$If!v h5}5f5h5f5g5g5e5e5 g5 e5 g5 e5 g#v}#vf#vh#vf#vg#ve#v g#v e#v g#v e#v g:V H!,,,,, , 55h5i5h5i5 h5 n9 / / 2 2 4 pւkd4$$If" n< p:m9!%%h$i%h%h$i$h%h%h%h%h$h$nH!44442 2 4apւ$$If!vh555#v#v#v:V 9!,,5R599/ /  2 24 p$$If!vh5555555#v#v#v#v#v#v:V H9!,5R559/  2 24 pF$$If!v h55q5q5p5p5p5r5p5 p5 p5 p5 p5 r#v#vq#vp#vr#v p#v r:V 9!,5R5 o9/ / 2 2 4 pւkd~;$$If" U6 hH*!%Roooooooooooo9!44442 2 4apւ$$If!v h55q5q5p5p5p5r5p5 p5 p5 p5 p5 r#v#vq#vp#vr#v p#v r:V 9!,, 5R5 o9 / / 2 2 4 pւkd?$$If" U6 hH*!%R$o$o$o$o$o$o$o$o$o$o$o$o9!44442 2 4apւ$$If!v h55q5q5p5p5p5r5p5 p5 p5 p5 p5 r#v#vq#vp#vr#v p#v r:V 9!,, 5R5 o9 / 2 2 4 pւkd:D$$If" U6 hH*!%R$o$o$o$o$o$o$o$o$o$o$o$o9!44442 2 4apւ$$If!v h55q5q5p5p5p5r5p5 p5 p5 p5 p5 r#v#vq#vp#vr#v p#v r:V 9!,, 5R5 o9 / 2 2 4 pւkdH$$If" U6 hH*!%R$o$o$o$o$o$o$o$o$o$o$o$o9!44442 2 4apւ$$If!v h55q5q5p5p5p5r5p5 p5 p5 p5 p5 r#v#vq#vp#vr#v p#v r:V 9!,, 5R5 o9 / 2 2 4 pւkdL$$If" U6 hH*!%R$o$o$o$o$o$o$o$o$o$o$o$o9!44442 2 4apւ$$If!v h55q5q5p5p5p5r5p5 p5 p5 p5 p5 r#v#vq#vp#vr#v p#v r:V 9!,, 5R5 o9 / 2 2 4 pւkd3Q$$If" U6 hH*!%R$o$o$o$o$o$o$o$o$o$o$o$o9!44442 2 4apւ$$If!v h55q5q5p5p5p5r5p5 p5 p5 p5 p5 r#v#vq#vp#vr#v p#v r:V 9!,, 5R5 o9 / 2 2 4 pւkdU$$If" U6 hH*!%R$o$o$o$o$o$o$o$o$o$o$o$o9!44442 2 4apւ$$If!v h55q5q5p5p5p5r5p5 p5 p5 p5 p5 r#v#vq#vp#vr#v p#v r:V 9!,, 5R5 o9 / 2 2 4 pւkdY$$If" U6 hH*!%R$o$o$o$o$o$o$o$o$o$o$o$o9!44442 2 4apւ$$If!v h55q5q5p5p5p5r5p5 p5 p5 p5 p5 r#v#vq#vp#vr#v p#v r:V 9!,,, , , 5R5 o9 / / 2 2 4 pւkd,^$$If" U6 hH*!%R%o%o%o%o$o%o%o%o%o%o$o%o9!44442 2 4apւ$$If!vh55E 5? #v#vE #v? :V  9!,5559/  / / / 2 24 pH$$If!vh55n5k5l5k5k5i#v#vn#vk#vl#vk#vi:V  9!,,5599/  / / / 2 24 pF$$If!v h5575756555656555 65 65 55 65 3#v#v7#v6#v5#v6#v5#v 6#v 5#v 6#v 3:V 9!,55M5L5M5L5 M5 L5 M5 L9/ / 2 2 4 pւkdd$$If"  J  V*!%MMMLMMLMMLML9!44442 2 4apւ$$If!v h5575756555656555 65 65 55 65 3#v#v7#v6#v5#v6#v5#v 6#v 5#v 6#v 3:V 9!, 55M5L5M5L5 M5 L5 M5 L9 / 2 2 4 pւkdi$$If"  J  V*!%%M%M%M%L%M%M%L%M%M%L%M%L9!44442 2 4apւ$$If!v h5575756555656555 65 65 55 65 3#v#v7#v6#v5#v6#v5#v 6#v 5#v 6#v 3:V 9!, 55M5L5M5L5 M5 L5 M5 L9 / 2 2 4 pւkdn$$If"  J  V*!%%M%M%M%L%M%M%L%M%M%L%M%L9!44442 2 4apւ$$If!v h5575756555656555 65 65 55 65 3#v#v7#v6#v5#v6#v5#v 6#v 5#v 6#v 3:V 9!, 55M5L5M5L5 M5 L5 M5 L9 / 2 2 4 pւkdr$$If"  J  V*!%%M%M%M%L%M%M%L%M%M%L%M%L9!44442 2 4apւ$$If!v h5575756555656555 65 65 55 65 3#v#v7#v6#v5#v6#v5#v 6#v 5#v 6#v 3:V 9!, 55M5L5M5L5 M5 L5 M5 L9 / / 2 2 4 pւkdWw$$If"  J  V*!%%M%M%M%L%M%M%L%M%M%L%M%L9!44442 2 4apւ$$If!vh59!#v9!:V 4  t 9!5"22q44 4Taf4p |$$If!vh59!#v9!:V 4 t9!5"/ 244 4Taf4$$If!vh55#v#v:V 9!559/ /  2 24 p$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd}$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / / 2 2 4 pւkdN$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkdސ$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd`$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd!$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkdd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd%$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkdh$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / / 2 2 4 pւkd)$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!vh55#v#v:V 9!559/ /  2 24 p$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / / 2 2 4 pւkd_$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / / 2 2 4 pւkd.$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd@$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkdD$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkd$$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / 2 2 4 pւkdH $$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!v h55I5K5J5L5K5L5K5 n5 K5 L5 K5 N#v#vI#vK#vJ#vL#vK#vL#vK#v n#v K#v L#v K#v N:V 9!55L5N5M5N5M5N5M5 p5 M5 N5 M5 P9 / / 2 2 4 pւkd $$If"  ^ AE*!&&L&N&M&N&M&N&M&p&M&N&M&P9!44442 2 4apւ$$If!vh555#v#v#v:V 9!,,5Q559/ /  2 24 pL$$If!vh5555555#v#v#v#v#v#v#v:V 9!,,5Q55555599/  2 24 pF$$If!v h55q5p5p5r5p5p5p5 r5 p5 p5 p5 r#v#vq#vp#vr#vp#v r#v p#v r:V 9!,, 5Q5 o99 / / 2 2 4 gpւkd$$If" T4 hH*!%Qoooooooooo$oo9!44442 2 4agpւ$$If!v h55q5p5p5r5p5p5p5 r5 p5 p5 p5 r#v#vq#vp#vr#vp#v r#v p#v r:V 9!,, 5Q5 o9 / / 2 2 4 pւkd$$If" T4 hH*!%Q$o$o$o$o$o$o$o$o$o$o$o$o9!44442 2 4apւ$$If!v h55q5p5p5r5p5p5p5 r5 p5 p5 p5 r#v#vq#vp#vr#vp#v r#v p#v r:V 9!,, 5Q5 o9 / 2 2 4 pւkd!$$If" T4 hH*!%Q$o$o$o$o$o$o$o$o$o$o$o$o9!44442 2 4apւ$$If!v h55q5p5p5r5p5p5p5 r5 p5 p5 p5 r#v#vq#vp#vr#vp#v r#v p#v r:V 9!,, 5Q5 o9 / 2 2 4 pւkdO&$$If" T4 hH*!%Q$o$o$o$o$o$o$o$o$o$o$o$o9!44442 2 4apւ$$If!v h55q5p5p5r5p5p5p5 r5 p5 p5 p5 r#v#vq#vp#vr#vp#v r#v p#v r:V 9!, 5Q5 o9 /  / / / 2 2 4 pւkd*$$If" T4 hH*!%Q%o%o%o%o%o%o%o%o%o%o%o%o9!44442 2 4apւ$$If!vh5B5 5 5#vB#v #v:V 9!5D5509/ /  2 24 p($$If!vh5B5555555#vB#v#v:V 9!5D5.509/  2 24 pP$$If!vh5B5555555#vB#v#v:V 9!5D5.509/  / 2 24 pP($$If!vh5B5555555#vB#v#v:V 9!5D5.509/ / 2 24 gpP$$If!vh5B5555555#vB#v#v:V 9!5D5.509/ 2 24 gpP$$If!vh5B5555555#vB#v#v:V 9!5D5.509/ 2 24 gpP$$If!vh5B5555555#vB#v#v:V 9!5D5.509/ 2 24 gpP$$If!vh5B5555555#vB#v#v:V 9!5D5.509/ 2 24 gpP$$If!vh5B5555555#vB#v#v:V 9!5D5.509/  2 24 gpP$$If!vh5B5555555#vB#v#v:V 9!5D5.509/  2 24 pP$$If!vh5B5555555#vB#v#v:V 9!5D5.509/  / 2 24 pP$$If!vh55751#v#v7#v1:V 6,,55V5S9/ /  2 24 p3$$If!vh5555555#v#v#v#v#v:V 6,,5555599/  2 24 pF$$If!v h55^5_5]5a5]5_5]5 _5 ]5 _5 ]5 \#v#v^#v_#v]#va#v]#v_#v]#v _#v ]#v _#v ]#v \:V 6,, 55c5d5c5e5c5d5c5 d5 c5 d5 c5 b99 / / 2 2 4 pւkd1=$$If" 5 Rk'?!%cdcecdcdcd$cb644442 2 4apւ$$If!vh5N!#vN!:V 6,59/ / 2 24 gp +$$If!v h55^5_5]5a5]5_5]5 _5 ]5 _5 ]5 \#v#v^#v_#v]#va#v]#v_#v]#v _#v ]#v _#v ]#v \:V 6,,,, , , 55c5d5c5e5c5d5c5 d5 c5 d5 c5 b9 / 2 2 4 gpւkdB$$If" 5 Rk'?!%$c$d$c$e%c$d$c$d$c$d%c$b644442 2 4agpւ+$$If!v h55^5_5]5a5]5_5]5 _5 ]5 _5 ]5 \#v#v^#v_#v]#va#v]#v_#v]#v _#v ]#v _#v ]#v \:V 6,,,, , , 55c5d5c5e5c5d5c5 d5 c5 d5 c5 b9 / 2 2 4 gpւkdG$$If" 5 Rk'?!%$c$d$c$e%c$d$c$d$c$d%c$b644442 2 4agpւ!$$If!v h55^5_5]5a5]5_5]5 _5 ]5 _5 ]5 \#v#v^#v_#v]#va#v]#v_#v]#v _#v ]#v _#v ]#v \:V 6,,,, , , 55c5d5c5e5c5d5c5 d5 c5 d5 c5 b9 / 2 2 4 pւkdL$$If" 5 Rk'?!%$c$d$c$e%c$d$c$d$c$d%c$b644442 2 4apւS$$If!v h55^5_5]5a5]5_5]5 _5 ]5 _5 ]5 \#v#v^#v_#v]#va#v]#v_#v]#v _#v ]#v _#v ]#v \:V 6,,,,,,,, , , , , 55c5d5c5e5c5d5c5 d5 c5 d5 c5 b9 / / 2 2 4 pւkdQ$$If" 5 Rk'?!%%c$d%c$e%c$d%c$d%c$d%c$b644442 2 4apւ$$If!vh5N!#vN!:V 6,59/ 2 24 gp O$$If!v h55^5_5]5a5]5_5]5 _5 ]5 _5 ]5 \#v#v^#v_#v]#va#v]#v_#v]#v _#v ]#v _#v ]#v \:V 6,,,,,,,, , , , , 55c5d5c5e5c5d5c5 d5 c5 d5 c5 b9 / 2 2 4 gpւkdrW$$If" 5 Rk'?!%%c$d%c$e%c$d%c$d%c$d%c$b644442 2 4agpւO$$If!v h55^5_5]5a5]5_5]5 _5 ]5 _5 ]5 \#v#v^#v_#v]#va#v]#v_#v]#v _#v ]#v _#v ]#v \:V 6,,,,,,,, , , , , 55c5d5c5e5c5d5c5 d5 c5 d5 c5 b9 / 2 2 4 gpւkd\$$If" 5 Rk'?!%%c$d%c$e%c$d%c$d%c$d%c$b644442 2 4agpւE$$If!v h55^5_5]5a5]5_5]5 _5 ]5 _5 ]5 \#v#v^#v_#v]#va#v]#v_#v]#v _#v ]#v _#v ]#v \:V 6,,,,,,,, , , , , 55c5d5c5e5c5d5c5 d5 c5 d5 c5 b9 / 2 2 4 pւkda$$If" 5 Rk'?!%%c$d%c$e%c$d%c$d%c$d%c$b644442 2 4apւS$$If!v h55^5_5]5a5]5_5]5 _5 ]5 _5 ]5 \#v#v^#v_#v]#va#v]#v_#v]#v _#v ]#v _#v ]#v \:V 6,,,,,,,, , , , , 55c5d5c5e5c5d5c5 d5 c5 d5 c5 b9 / / 2 2 4 pւkdf$$If" 5 Rk'?!%%c$d%c$e%c$d%c$d%c$d%c$b644442 2 4apւo$$If!vh5O5N5N5N#vO#vN:V 9!5/ 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 / 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 / 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 4$$If!v h5O55555555 5 #vO#v#v#v#v#v#v :V 9!5555555 / 4$$If!vh5@ 5 5* #v@ #v #v* :V 9!555d9/ /  2 24 p$$If!vh5@ 55555#v@ #v#v#v:V 9!55559/  /  2 24 p<$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / / 2 2 4 pngkdx$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / / 2 2 4 pngkd}$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd $$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdۨ$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd֬$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdѰ$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd̴$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdǸ$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd¼$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd|$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdw$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdr$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdm$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdh$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdc$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkd^ $$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdY$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdT$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdO$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / 2 2 4 pngkdJ$$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!v h5@ 5L5M5M5N5M5N5M5 N5 M5 B#v@ #vL#vM#vN#vM#vN#vM#v N#v M#v B:V 9!55Z5[5Z5 [5 Z5 U9 / / 2 2 4 pngkdE $$If 1 } eM*!$$Z$Z$Z$Z$Z$[$Z$[$Z$U9!,,,,2 2 4apn$$If!vh5 5 5 #v #v #v :V 49!5-559/ / 2 24 p+$$If!vh5 575959575759#v #v7#v9#v7#v9:V 49!5-55559/  / 2 24 pF($$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ / 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/  2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/  2 24 pF$$If!vh5 5959575759#v #v9#v7#v9:V 9!5 5559/ 2 24 p<$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pFv$$If!vh59!#v9!:V 9!59/ 2 24 p $$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/ 2 24 pF$$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/  2 24 pF($$If!vh5 575959575759#v #v7#v9#v7#v9:V 9!5-55559/  / 2 24 pF$$If!vh5 5: 57 #v #v: #v7 :V 9!5559/ / 2 24 p<$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/  / 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF$$If!vh5/5i5j5g5i5g#v/#vi#vj#vg#vi#vg:V 9!5 55559/ 2 24 p<.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/ 2 24 pF.$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/  2 24 pF<$$If!vh5 5g5i5j5g5i5g#v #vg#vi#vj#vg#vi#vg:V 9!5555559/  / 2 24 pF$$If!vh5 5z5}5}#v #vz#v}:V 9!5559/ / 2 24 p(d$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / / 2 2 4 pd7kdz$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdV$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / 2 2 4 pd7kd-~$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdV$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / 2 2 4 pd7kd$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdV$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / 2 2 4 pd7kdO$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdV$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / 2 2 4 pd7kd$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdV$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / 2 2 4 pd7kdq$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdV$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / 2 2 4 pd7kd$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdV$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / 2 2 4 pd7kd$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdV$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / 2 2 4 pd7kd$$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdV$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / 2 2 4 pd7kd$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdV$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / 2 2 4 pd7kdF$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apdd$$If!v h5 5(5)5)5)5+5)5)5 )5 +#v #v(#v)#v+#v )#v +:V 9!55E5F5 E5 F9 / / 2 2 4 pd7kdס$$If 0Y*!$$E$E$E$E$F$E$E$E$F9!((((2 2 4apd$$Ifq!vh55a5a5a5a#v#va:V q9!559/ /  22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/ 22 q4 qaqp2$$Ifq!vh55a5a5a5a#v#va:V q9!559/  / 22 q4 qaqp2$$Ifl!vh5|5 #v|#v :V l9!5 5}9/ /  22 lalp$$Ifl!vh5|5 #v|#v :V l9!5 5}9/ / 22 lalp$$Ifl!vh5|5 #v|#v :V l9!5 5}9/ 22 lalp$$Ifl!vh5|5 #v|#v :V l9!5 5}9/ 22 lalp$$Ifl!vh5|5 #v|#v :V l9!5 5}9/  22 lalp$$Ifl!vh5|5 #v|#v :V l9!5 5}9/  / 22 lalp$$Ifq!vh555#v#v:V q49!5 5k9/ /  22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ / 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/  22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/  22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/  22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/  22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/  22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/ 22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/  22 q4 qaqp$$Ifq!vh555#v#v:V q9!5 5k9/  / 22 q4 qaqp^G pp02 0@P`p2( 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p8XV~_HmH nH sH tH N`N sNormal dCJOJQJ_HmH sH tH t`t 7s Heading 1'$$<&d@&Pa$ CJ<KHOJQJ_HmH sH tH T`T .g Heading 2 $@&CJ$OJQJ_HmH sH tH R`R s Heading 3 @&CJOJQJ_HmH sH tH V V s Heading 4 @&6CJOJQJ_HmH sH tH @@ m Heading 5$@& 5OJQJHH  Heading 6$ @&^ CJ$OJQJNN  Heading 7$$@&a$5@CJ$OJQJPP  Heading 8$$@&a$5@<B*CJ8OJQJDA`D sDefault Paragraph FontVi@V  Table Normal :V 44 la (k (sNo List JoJ /8+Text dCJOJQJ_HmH sH tH >)@> s Page NumberCJOJQJkHJ @J sFooter 9!dCJOJQJtH Z`"Z spTOC 1 9!P^CJOJQJ_HmH sH tH F@F spTOC 2 9! ^ CJ.. TOC 3 ^hoh <s tabletitleShP^S`5CJOJQJ_HmH sH tH PoP s Table textPCJOJQJ_HmH sH tH .. TOC 4 ^.. TOC 5 p^p.. TOC 6 L^L.. TOC 7 (^(.. TOC 8 ^.. TOC 9 ^XoX s Tablehead1 PP5CJOJQJ_HmH sH tH ll sColorful Grid - Accent 1 7P]^7CJXoX s References^`OJQJ_HmH sH tH DOD s Tablehead2 54O4 s Tablehead3!6X#@X s0Table of Figures" 9! WP^ `WN2N sHeader# 9!dCJOJQJtH <OB< NImprint$d CJOJQJPQRP s Figuretitle%$ S^S`tH lobl sPublication Title &$ a$$CJ<KHOJQJ_HaJ<mH sH tH VorV sAuthors'$dpa$CJ OJQJ_HmH sH tH lol M Dotpoint1$( & F hx^`CJOJQJ_HmH sH tH R/R s Dotpoint2 ) & FCJOJQJ_HmH sH tH h/h a9contents*$&dPa$ CJ<KHOJQJ_HmH sH tH / fSNumberedListContinuing+ & Fx^`CJOJQJ_HmH sH tH XoX ;sSource,7(^7`CJOJQJ_HmH sH tH h@h :s Footnote Text$- Vd$^`VCJOJQJtH 4U4 eE Hyperlink >*phF/F 8+ Text CharCJOJQJ_HmH sH tH L&@L sFootnote ReferenceCJH*OJQJB'B `Comment ReferenceCJaJ8"8 3` Comment Text2CJJ1J 2`Comment Text Char_HmH sH tH HBH ` Balloon Text4CJOJQJ^JaJ@j!"@ 6Comment Subject55\F2aF 5Comment Subject Char5\T/qT ]Heading 1 Char CJ<KHOJQJ_HmH sH tH \\ zgrezidascomments008B*aJmH phsH tH JoJ stext 9dCJOJQJ_HmH sH tH POP -umnFootnote Text CharCJOJQJmH sH J/J ,*K Source CharCJOJQJ_HmH sH tH V/V /|tabletitle Char5CJOJQJ_HmH sH tH ll > Colorful List - Accent 1=^CJOJPJQJaJJBJ ?< Body Text>5CJmH nHsH tHHH ><Body Text Char5mH nHsH tHJPJ A< Body Text 2@CJmH nHsH tHLL @<Body Text 2 CharCJmH nHsH tHV/"V Default B7$8$H$!B*CJ_HaJmH phsH tH fo2f s institutionC$da$#:CJOJQJ_HaJmH sH tH @OB@ stext-less#beforeDP8OR8 s text-more#b4Ehjcj ^w Table Grid7:VF0FPK![Content_Types].xmlj0Eжr(΢Iw},-j4 wP-t#bΙ{UTU^hd}㨫)*1P' ^W0)T9<l#$yi};~@(Hu* Dנz/0ǰ $ X3aZ,D0j~3߶b~i>3\`?/[G\!-Rk.sԻ..a濭?PK!֧6 _rels/.relsj0 }Q%v/C/}(h"O = C?hv=Ʌ%[xp{۵_Pѣ<1H0ORBdJE4b$q_6LR7`0̞O,En7Lib/SeеPK!kytheme/theme/themeManager.xml M @}w7c(EbˮCAǠҟ7՛K Y, e.|,H,lxɴIsQ}#Ր ֵ+!,^$j=GW)E+& 8PK!Ptheme/theme/theme1.xmlYOo6w toc'vuر-MniP@I}úama[إ4:lЯGRX^6؊>$ !)O^rC$y@/yH*񄴽)޵߻UDb`}"qۋJחX^)I`nEp)liV[]1M<OP6r=zgbIguSebORD۫qu gZo~ٺlAplxpT0+[}`jzAV2Fi@qv֬5\|ʜ̭NleXdsjcs7f W+Ն7`g ȘJj|h(KD- dXiJ؇(x$( :;˹! I_TS 1?E??ZBΪmU/?~xY'y5g&΋/ɋ>GMGeD3Vq%'#q$8K)fw9:ĵ x}rxwr:\TZaG*y8IjbRc|XŻǿI u3KGnD1NIBs RuK>V.EL+M2#'fi ~V vl{u8zH *:(W☕ ~JTe\O*tHGHY}KNP*ݾ˦TѼ9/#A7qZ$*c?qUnwN%Oi4 =3ڗP 1Pm \\9Mؓ2aD];Yt\[x]}Wr|]g- eW )6-rCSj id DЇAΜIqbJ#x꺃 6k#ASh&ʌt(Q%p%m&]caSl=X\P1Mh9MVdDAaVB[݈fJíP|8 քAV^f Hn- "d>znNJ ة>b&2vKyϼD:,AGm\nziÙ.uχYC6OMf3or$5NHT[XF64T,ќM0E)`#5XY`פ;%1U٥m;R>QD DcpU'&LE/pm%]8firS4d 7y\`JnίI R3U~7+׸#m qBiDi*L69mY&iHE=(K&N!V.KeLDĕ{D vEꦚdeNƟe(MN9ߜR6&3(a/DUz<{ˊYȳV)9Z[4^n5!J?Q3eBoCM m<.vpIYfZY_p[=al-Y}Nc͙ŋ4vfavl'SA8|*u{-ߟ0%M07%<ҍPK! ѐ'theme/theme/_rels/themeManager.xml.relsM 0wooӺ&݈Э5 6?$Q ,.aic21h:qm@RN;d`o7gK(M&$R(.1r'JЊT8V"AȻHu}|$b{P8g/]QAsم(#L[PK-![Content_Types].xmlPK-!֧6 +_rels/.relsPK-!kytheme/theme/themeManager.xmlPK-!Ptheme/theme/theme1.xmlPK-! ѐ' theme/theme/_rels/themeManager.xml.relsPK] 3S4r:CVZFqCRjF- l^) M tCtPPPPP= J^^^a!=r!#&;Vlb;A=>I@wAYCDFGeSn;hŏ   '2@`=EXv-\,bd6gǖ[[fOͫhi;ž QY/E3giXDESTwx$%CDjk8;0+7Ok/Kdr;:.;5;;V<<==>M??e@@xAABECCWDEEFF^G N_``K`Y``````a:acaiaaaabDblbbbbbbccLc[ccccccd%d9d_dqddddd e/eieeeeeffffff7f=fFftfffffg4gMgsgggggg#h6hbhhhhhhh i7iciwiiijj5j[j|jjjkkPkukzkkkkkkl&ljlll@mmmmm nQn|})}T}x}~}}}}}}~~/~C~_~~~~~~~+G`}11>$0B ,6CPcmz 7Ob8Wt>nVuZw6K[rVey &Si~*h_Er%l)[F B8ŏ   !"#$%&()*+,-./013456789:;<=>?ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<>?<T /7SUi8;[wz36Wsv:={ 9<Rnq+[wzMil/2Xtw1MP!=@{X %%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%% QX[a!!  ?2$YeŌ,uBTtb$8Vн Űf2$G,x]mEOSe@ H 0(  : (  t =Ue(;  #  C"?BB     e(Ue(;BB  B   =;e(;   C BA*? ncver right tab_mono" 8Ue( `  A?Öbn-DEEWR1APicture 8n-DEEWR1A.eps h'pIH`'//C8PFQ&Q'Q=Qy>PF`T@`T`'q?E h h3" ?h =Ue(; #  #" ?PB    "e(Ue(;PB B   ""=;e(;d  C A? "&8Ue( B S  ?K  g#>6tdM t#l2t@N OLE_LINK3 _Toc495748330 _Toc495810630 _Toc6031787 _Toc6031844 _Toc256606224 _Toc148784740 _Toc150413029 _Toc151540209 _Toc148685435 _Toc148784741 _Toc150413030 _Toc151540210 _Toc256682419 _Toc260658316 _Toc260759990 _Toc268097912 _Toc274728186 OLE_LINK4 _Toc47346280 _Toc47951941 _Toc107393567 _Toc256682420 _Toc260658317 _Toc260759991 _Toc274728187 _Toc151540211 _Toc457122463 _Toc151540212 _Toc151540213 _Toc151540214 _Toc151540215 _Toc151540216 _Toc151540217 _Toc151540218 OLE_LINK1 OLE_LINK2 _Toc151540219 _Toc151540220 _Toc151540221 _Toc151540222 _Toc151540223 _Toc153533899 _Toc151540224 _Toc153533900 _Toc153533901 _Toc151540260 _Toc153533902 _Toc153533903 _Toc153533904 _Toc151540225 _Toc153533905 _Toc153533906 _Toc153533907 _Toc151540226 _Toc153533908 _Toc151540227 _Toc151540228 _Toc151540229 _Toc151540230 _Toc151540231 _Toc151540232 _Toc153533909 _Toc151540233 _Toc153533910 _Toc153533911 _Toc151540234 _Toc153533912 _Toc456000800 _Toc457122465 _Toc151540235 _Toc151540236 _Toc151540237 _Toc153533913 _Toc153533914 _Toc153533915 _Toc268097939 _Toc274728213 * * * * * } +(3JGO4V4V\]m[8;;Xb) D&E+N2OUq9æ`KT@  !"#$%&'()*+,-./0123456789:;<=>?@ABCJKDEFGHILM ) K K K K K ~ *5((3]GOzVzV\]mgg vhtE &j+2PUqLΦjS&&&&Vmostx7Vi<[{7Ww>{ =Rr[{Mm3Xx1Q!A{&'#.  "#%&()*5ny{|&*5ny7Vi<[{7Ww>{ =Rr[{Mm3Xx1Q!A{&  *5nyhB+(y>H@hZ J2#$ voLj2 ~H*9(M'q,J) Qz~>*ЬWRZ)x`(4ymEr*xp(k^`}.( ^`o(hH ^`hH. pLp^p`LhH. @ @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PLP^P`LhH.hh^h`B*OJQJo(h^`CJOJQJaJo(hHh^`OJ QJ ^J o(hHohpp^p`OJQJo(hHh@ @ ^@ `OJQJo(hHh^`OJ QJ ^J o(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJ QJ ^J o(hHohPP^P`OJQJo(hHh  ^ `OJQJo(hHh ^`hH.h L^`LhH.h | | ^| `hH.h LL^L`hH.h L^`LhH.h ^`hH.h ^`hH.h L^`LhH.^`OJQJo(hH ^`hH. pp^p`hH. @ @ ^@ `hH. ^`hH. ^`hH. ^`hH. ^`hH. PP^P`hH.h ^`o(hH.h^`OJ QJ ^J o(hHohp^p`OJQJo(hHh @ ^@ `o(hH.h^`OJ QJ ^J o(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJ QJ ^J o(hHohP^P`OJQJo(hHh ^`o(hH.h^`OJ QJ ^J o(hHohp^p`OJQJo(hHh@ ^@ `OJQJo(hHh^`OJ QJ ^J o(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJ QJ ^J o(hHohP^P`OJQJo(hHh ^`hH.h ^`hH.h pL^p`LhH.h @ ^@ `hH.h ^`hH.h L^`LhH.h ^`hH.h ^`hH.h PL^P`LhH.h^`CJOJQJaJo(hHh^`OJ QJ ^J o(hHohpp^p`OJQJo(hHh@ @ ^@ `OJQJo(hHh^`OJ QJ ^J o(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJ QJ ^J o(hHohPP^P`OJQJo(hHh^h`B*OJQJo(wh^`OJQJo(hHh^`OJ QJ ^J o(hHohp^p`OJQJo(hHh @ ^@ `o(hH.h^`OJ QJ ^J o(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJ QJ ^J o(hHohP^P`OJQJo(hHh ^`hH.h ^`hH.h pLp^p`LhH.h @ @ ^@ `hH.h ^`hH.h L^`LhH.h ^`hH.h ^`hH.h PLP^P`LhH.h ^`hH.h ^`hH.h pLp^p`LhH.h @ @ ^@ `hH.h ^`hH.h L^`LhH.h ^`hH.h ^`hH.h PLP^P`LhH. ^`hH. ^`hH. pp^p`hH. @ @ ^@ `hH. ^`hH. ^`hH. ^`hH. ^`hH. PP^P`hH.h^`OJQJo(hHh^`OJ QJ ^J o(hHohp^p`OJQJo(hHh @ ^@ `o(hH.h^`OJ QJ ^J o(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJ QJ ^J o(hHohP^P`OJQJo(hHh ^`hH.h ^`hH.h pLp^p`LhH.h @ @ ^@ `hH.h ^`hH.h L^`LhH.h ^`hH.h ^`hH.h PLP^P`LhH.h^`OJQJo(hHh ^`hH.h pLp^p`LhH.h @ @ ^@ `hH.h ^`hH.h L^`LhH.h ^`hH.h ^`hH.h PLP^P`LhH.h   ^ `hH.h ^`hH.h L^`LhH.h | | ^| `hH.h LL^L`hH.h L^`LhH.h ^`hH.h ^`hH.h L^`LhH.q,J~>^`} L#$ ym Q*9r)x`\ WRZ \ Z *x>Hq,JЕ,ӕ,a1U0V a1=$        >8V                                                     >8V                                                                                   .J `^ XuBTy$Q*]Fn/#UL& ]%,&o_6+X&,^,e6%q; ; '?9@C\uL'_L?N}/N;PrQT]VW VAW^]+_yH`dTbb:k 0mr Gmj o8tEUjxE6z^UzZ{s[2|/W_}')@d3S3T3b3c@\@l@@ @UnknownAuthor G* Times New Roman5Symbol3. * Arial]& zAvant GardeCentury Gothic7K@Cambria7.{ @Calibri9Garamond5. *aTahoma;Wingdings?= * Courier NewA BCambria Math#h:!G:!GF+ZF+Z$4d66 2qX ?G6! xx T             Oh+'0d    , 8DLT\  Normal.dotm1Microsoft Office Word@@do@= @= +F՜.+,D՜.+, `hpx  Z6  Title 8@ _PID_HLINKSAlKhttp://www.voced.edu.au/  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]_`abcdefghijklmnopqrstuvwxyz{|}~Root Entry F2= Data 1Table^@WordDocument SummaryInformation(DocumentSummaryInformation8CompObjy  F'Microsoft Office Word 97-2003 Document MSWordDocWord.Document.89q