Understanding the rural - urban reading gap - November 2002

4. What rural characteristics lie behind the difference in reading performance?

Previous

Contents

Next

The difference in some provinces between the performance of students in rural and urban schools is not related to differences in the schools (which are few) but rather to differences in the nature of work found in rural and urban communities.

The first temptation when comparing student performance from different regions is to look to the schools to find out what it is about the education system that can explain differing results. The initial analysis of rural-urban differences in this report demonstrated, however, that the data collected in PISA show that while there are some differences between rural and urban schools, they do not reflect the rural-urban reading differences.

There were, however, a number of variables that differed significantly between rural and urban communities. An examination of these variables showed that, in fact, the rural-urban difference in reading performance that exists in some provinces is best explained by differences in the kinds of jobs in the communities where these schools were located, specifically in the education-level required by those jobs.

In order to determine which factors best explain the difference between the reading performance of rural and urban students, the actual reading averages of rural students in each province were compared with the averages they would have attained if there were no rural-urban difference in a variety of important individual, family, school and community factors. For example, after estimating the relationship between parental education and individual performance, it is possible to calculate the expected individual performance corresponding to various levels of parental education. Using this technique, this analysis calculates what the average performance of students in rural areas would be if their families, schools, and communities had the average characteristics of the urban areas in their provinces.

While individual and family characteristics explain some of the rural-urban difference in student reading performance, their explanatory value is small.

The first phase in this analysis looked at the relationship between a combination of important individual and family factors and the rural-urban gap. The results, shown in table B1, indicate that while individual and family characteristics explain some of the rural-urban difference in student reading performance, their explanatory value is small. When the differences in individual and family factors are taken into account, there remains a large, systematic difference between rural and urban students that still needs to be explained. That is, controlling for the family background of individual students, rural students still show lower reading scores.

Column 1 shows the size of the rural-urban gap between the actual reading performance of students in rural and urban schools (a negative number means that rural student performance is lower by the number indicated). Column 2 shows the size of the difference that remains after controlling for individual students' socioeconomic status — that is, the difference between actual reading performance of urban students and the expected performance of rural students if they had the same family socio-economic characteristics as their urban counterparts (parental education and occupational status at the individual level).

In the provinces where there is a substantial difference in the performance of students in rural and urban schools, only a small part of this difference is explained by the differences in socio-economic status at the student level. This means that one cannot attribute the rural-urban difference in performance solely to differences in the socio-economic status of individual students. Urban students would still be expected to perform better than rural students even if their parents had the same education and similar jobs. In spite of the strong effect of a student's family socioeconomic status on reading performance found in the first results of PISA, there are other important factors at play in explaining the difference in rural and urban results.

Measures of family and community background

This analysis examined a variety of variables related to family background (socio-economic status) of students, as well as the characteristics of the communities where their schools were located. Variables describing the socio-economic background of students' families were the highest occupational status and the highest level of education of students' parents. The occupational status of parents was derived using the International Socio-economic Index of Occupational Status (see Appendix B for definition).

As measures of the community environment of students, the analysis examined average occupational and educational status of the parents of 15-year-olds in the school, average income, employment rates, unemployment rates, the proportion of adults with postsecondary education, and specifically with university education, and the proportion of adults working in jobs that generally require university qualifications (white collar jobs).

The second phase in this analysis looked at the relationship between a variety of community factors and the remaining rural-urban gap. To do this, a model was developed that looked at the individual and combined effect of a variety of important community variables such as community employment rates, the average educational attainment of the adult population, and the average education and occupational status of all the parents of the 15-year-olds in the school. This model confirmed that differences between rural and urban communities best explain the differences in rural and urban reading performance. More specifically, community characteristics related to adult education levels have the strongest relationship with the rural-urban gap.

Differences between rural and urban communities best explain the differences in rural and urban reading performance.

Column 3 of table B1 shows the remaining difference after controlling for these community variables as well as individual socio-economic status (parental occupation and educational attainment). In all provinces, except Alberta, the rural-urban difference disappeared completely once community variables related to adult educational attainment were controlled for. In fact, after controlling for the community level variables, rural students were actually expected to outperform students from urban schools. This means that if the rural communities had these same characteristics as urban communities in the same province, rural students would be expected to score higher than urban students.

In all provinces, the relative impact of these variables is considerable. A comparison of the difference that remains after controlling for individual socio-economic status and the difference after further controlling for the community variables reveals that the community factors far outweigh individual student socio-economic status in accounting for the difference in rural and urban student performance. For example, in Alberta, where there is an actual difference of 21 points between rural and urban students, when individual socio-economic status is controlled for, there is still a difference of 17 points. However, when the combination of community variables is taken into account, this difference shrinks to only 5 points.

Figure 5
Controlling for the difference in family and community background, students in rural schools would actually outperform their urban counterparts in most provinces

The average occupational status of the parents of the 15-year-olds in the school is the factor with the strongest relationship with the rural-urban difference in reading performance.

An examination of the variables that contribute to this model reveals that the average occupational status of the parents of the 15-year-olds in the school is the factor with the strongest relationship with the rural-urban difference in reading performance. Columns 4 to 10 show the size of the differences after controlling for the impact of the community factors one at a time. In all provinces, when just aggregate parental occupational status is controlled for, rural students perform about as well as or better than urban students. Other important community characteristics that contribute to the rural-urban gap are the average educational attainment of adults, the proportion of adults with postsecondary education (and more specifically university education), and the proportion of workers whose jobs require university training. Community average employment rates and average income had smaller or larger effects depending on the province. In general, these variables are all related to the education-level of jobs in the community. While it is unlikely that any of these variables directly causes lower student achievement, they are indicators of the environment in which these students learn and look for support for their learning. In general, they may be a visible measure of the degree to which higher education pays off in the community.

Further analysis is required for a full understanding of the relationship between these community characteristics and student performance. To what extent do education levels of adults in the community reflect the availability of role models who demonstrate the value of education? To what extent do they reflect the availability of career options that require further education? These questions and others have yet to be answered. This study was intended as a means of investigating the rural-urban difference with a variety of variables available from YITS/PISA. Other variables available both from the YITS/PISA database can provide more detailed information about how the community level variables interact with students education and career expectations, family background and parental expectations, as well as peer influence. Furthermore, information collected in 2002 when these students were 17 years old will provide an additional set of information on factors influencing decisions about education and labour market choices.

Previous

Contents

Next