Chapter 2
Data and methods

View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Archived

This page has been archived on the Web.

[an error occurred while processing this directive]81-595-m[an error occurred while processing this directive] [an error occurred while processing this directive]

2.1 Data

The Youth in Transition Survey (YITS) is a Canadian longitudinal survey designed to examine the patterns of, and influences on, major transitions in young people's lives, particularly with respect to education, training and work. In order to address these objectives, data were collected from two cohorts of youth in the first cycle of the survey in 2000. One cohort began its participation at age 15 (Cohort A) and the other at ages 18 to 20 (Cohort B). In this report, all five cycles of Cohort B are used, providing information every two years from 2000 to 2008. At the time of the last cycle in 2008, respondents were aged 26 to 28. All longitudinal surveys are subject to sample attrition. Initially in Cycle 1, over 22,000 cases were available; by Cycle 2, this number decreased to fewer than 19,000; by Cycle 3, it decreased further to fewer than 15,000. Cycle 4 saw a further decline to just less than 12,500, while in the final cycle of data available in 2008 9,946 cases were available for analysis. Since this report uses all five cycles, survey weights from Cycle 5 are used in order to maintain generalizability due to attrition.

2.2 Data restrictions and sample design

Data from YITS are used to assess labour market outcomes at two points in time following the exit from full-time schooling: Time 1 refers to 1 or 2 years after leaving school and Time 2 refers to 5 or 6 years after leaving school. To create the sample for analysis, the main restriction was with regard to date last in school full-time. To be able to analyze labour market outcomes up to six years after leaving school, it was necessary to only include respondents who left school in December 2002 or earlier. Additionally, a small number of youth who left school on a full time basis prior to January 1997 were also removed from the analysis. The result is a sample of greatly reduced size; however, these restrictions were necessary in order to have a sample with valid responses on labour market outcomes at two points in time after leaving full-time school. If the sample is not the same across time, it is difficult to disentangle the changes due to individual behaviour or changes due to sample composition. Other work has utilized a similar approach with success (see Finnie's 1999 work using two cycles of National Graduates Survey (NGS) graduates). These substantial sample restrictions notwithstanding, the benefits to this approach are two-fold: first, by focussing on the date of school leaving rather than the date of graduation, youth who have left the education system without any credentials can be included, and second, it allows for the study of how certain educational pathways can impact upon one's labour market trajectory over a six-year period.

After imposing the aforementioned selection criteria based on time when respondents left full-time school, as well as removing any missing information from covariates on a listwise basis, the sample of 9946 was reduced to 3592 for the employment estimations and 3042 for the earnings estimations. For the ordinary least squares regressions pertaining to earnings, the sample size is somewhat smaller because they are restricted to youth with positive earnings at each respective time point. The drawback to a restrictive sample such as this is that it is more likely to include youth with lower education levels and of lower socioeconomic status, since it excludes youth who were still attending school full time and who, therefore, were likely to be enrolled in either college or university.

Table 2.1, for instance, compares select sample characteristics for the main sample as well as the two samples used for analyzing labour market outcomes. In general, the samples restricted to youth who left school no later than December 2002 are more likely to have lower education than the total YITS sample since the earlier leaving dates are biased towards a younger age at leaving and lower education levels. For instance, the average age of youth who left full-time school in the original sample is almost 23, yet in the restricted samples they are about age 20. Moreover, the proportion of the original sample with a high school diploma only is approximately 9%, yet in the restricted samples, this same proportion is closer to 20%. Conversely, in the original sample, the proportion with a university degree (without having first completed a college diploma) is around 12%¹, while in the restricted samples the proportion is much lower (between 6.3% and 2.3%). We also observe from Table 2.1 that youth in the samples used for analyses are less likely to have parents who have completed post-secondary degrees / diplomas and are more likely to be Canadian born. This last statement reflects the fact that immigrants have higher education levels than the Canadian born, on average.

Table 2.1 Descriptive statistics by sample restrictions, weighted

---

Note

1. Approximate average of 14.59% for Non-gappers-University(+) and 9.38% for Gappers-University(+) is 11.99%.