Chapter 3: Differences in performance by selected student characteristics

View the most recent version.

Chapters >

Introduction

As shown in Chapter 1, the science performance of students differed considerably across countries and across provinces. In this Chapter, selected student characteristics that are related to performance are explored beginning with an examination of performance by gender and immigrant status followed by an examination analysis of the impact of parental education and of socio-economic status.

Boys outperform girls in 'explaining phenomena scientifically' while girls outperform boys in 'identifying scientific issues'
Immigrant students perform lower in science than non-immigrant students
Higher levels of parental education are associated with higher performance in science
Socioeconomic status is related to performance but to a lesser extent in Canada when compared to other countries
Summary

Boys outperform girls in 'explaining phenomena scientifically' while girls outperform boys in 'identifying scientific issues'

Policy makers have an interest in reducing gender disparities in educational performance. Such performance coupled with their motivation and attitudes towards learning influence both educational and occupational pathways of boys and girls.

PISA 2006 shows that gender differences on the combined science scale, which varied across countries, were small in absolute terms when compared with the large gender gap in reading. In Canada, no gender difference was observed on the combined science scale with the exception of Newfoundland and Labrador where girls performed better than boys by 12 score points. Across all countries participating in PISA 2006, ten countries showed an advantage of boys over girls while thirteen countries showed an advantage of girls over boys. For the remaining countries there were no significant gender differences on the combined science scale.

In Canada, although overall there were no gender differences on the combined science scale or on the sub-scale of using scientific evidence, there were substantial gender differences on the other two science sub-scales as summarized in Table 3.1. In Canada, most countries and eight of the ten provinces, boys out performed girls in the sub-domain of 'explaining phenomena scientifically'. Canadian boys outperformed girls by 17 score points and this difference ranged from 0 to 21 points across provinces (Appendix table B.3.1) while across all OECD countries boys outperformed girls by 15 score points. In contrast, in Canada, most countries and all of the provinces, girls outperformed boys in the sub-domain 'identifying scientific issues'. The magnitude of this difference was 14 points for Canada overall, 17 points across all OECD countries and ranged from 12 to 36 score points across provinces.

The performance patterns on these two sub-scales suggest that boys and girls have very different levels of performance in different areas of science. It appears that boys demonstrate better performance at mastering scientific knowledge whereas girls demonstrate better performance at seeing the larger picture that enables them to identify scientific questions that arise from a given situation.

Gender differences across the minor domains of mathematics and reading were consistent with results from previous PISA assessments with boys outperforming girls in mathematics and girls outperforming boys in reading. As was also the case in previous PISA assessments, the gap between girls and boys in reading was much larger than the gap between boys and girls in mathematics. In Canada, while boys outperformed girls by 14 points in mathematics, girls outperformed boys by 33 points in reading. At the provincial level, boys outperformed girls in mathematics in six of the ten provinces while girls outperformed boys in reading in all provinces.

When examining patterns of gender differences for the domains across provinces, some interesting patterns emerge. In contrast to that observed for Canada overall and the remaining provinces, in Newfoundland and Labrador, Prince Edward Island and Saskatchewan, no gender differences were observed in mathematics or on the sub-scale of 'explaining phenomena scientifically'. In three of the largest provinces, Quebec, Ontario and British Columbia, the magnitude of the gender differences across domains were more pronounced in domains where boys outperformed girls and less pronounced in domains where girls outperformed boys (Appendix tables B.3.1 and B.3.2).

Immigrant students perform lower in science than non-immigrant students

Immigration has long been integral to Canada's social, cultural and economic development and it is expected that immigrants will constitute a larger share of the growth of Canada's labour force in the future. Results of the 2003 International Adult Literacy and Skills Survey¹ revealed that a larger proportion of adult immigrants performed at a lower level in literacy, numeracy and problem solving than Canadian-born. PISA data can also be explored to see if performance differences exist between immigrant and non-immigrant students.

PISA identifies two groups of immigrants: second-generation immigrants are those born in Canada from immigrant parents; and first generation immigrants are those born outside of Canada. Students born in Canada from parents also born in Canada are defined as non-immigrant students.

Twenty one percent of the Canadian youth that participated in PISA 2006 were immigrants. Ten percent were first generation and 11% second generation. Immigrant youth represent 10% or more of the 15-year-old youth in five provinces: Quebec, Ontario, Manitoba, Alberta and British Columbia (Appendix table B.3.3). Detailed results about immigrant youth are presented only for these five provinces.

Second-generation immigrants completed all their education in Canada, having been born in this country. However this is not necessarily the case for first-generation immigrants depending on their age of arrival to Canada. Even if PISA is not measuring knowledge and skills specifically acquired in school, one would expect that the differences in scores between second-generation immigrants and non-immigrant students would be small, particularly in an area like science, where youth are more likely to develop their knowledge and skills in school rather than in their family or community. However, as shown in Table 3.2, non-immigrant students outperformed second generation immigrants who were born and educated in Canada. In three provinces (Manitoba, Quebec and British Columbia) the performance differences between non-immigrant youth and their second generation peers were significant. Within Canada, the difference ranged from 20 points in Manitoba to 39 points for Quebec.

A similar pattern is observed when looking at differences in performance between first generation immigrants and non-immigrant students. In Canada, non-immigrant youth outperformed their first-generation peers by 23 points in science. Performance differences were significant in Ontario (27 points) and Manitoba (32 points) and were most pronounced in Quebec at 57 points.

When comparing performance between first and second generation immigration youth, there were similar levels of performance in four provinces. However, for Canada overall and Ontario second generation immigrant students outperformed first generation immigrant students by 10 points and 19 points respectively.

It should be noted that immigrant youth in Canada have much higher performance than immigrants across all OECD countries (Appendix table B.3.3) and also perform above the OECD average of 500. In addition, the magnitude of performance differences of immigrant students compared to non-immigrant students is less pronounced in Canada compared to across all OECD countries as whole. The only exception is observed in Quebec where the difference in scores is at the OECD average. In contrast, in Alberta there were no differences in performance between immigrant and non-immigrant students.

Higher levels of parental education are associated with higher performance in science

Parents play an important role in how students learn. Aside from being actively involved in their children's education, parents also provide a home environment that can impact on learning. Parents serve as a model for learning, determine the educational resources available in the home and hold particular attitudes and values towards education.

Parental education is a factor that has been shown to be related to important transitions in youth life. For example, Knighton and Bussière (2006) found that youth whose parents highest level of education was high school or less were less likely to complete high school or to go to post-secondary education by age 19.

Thirty percent of the Canadian youth that participated in PISA 2006 had parents whose highest level of education was high school or less while 70% of students had a parent with at least some post secondary education. Parental education varied across provinces (Appendix table B.3.4):  students whose parents had some postsecondary education ranged from 56% (Newfoundland and Labrador) to 77% (Ontario).

Youth with at least one parent who had post-secondary education outperformed their peers whose parents had high school education or less (Chart 3.1). In Canada, this difference amounted to 32 score points which is roughly equivalent to one year of education. While differences existed across all of the provinces the magnitude of this difference varied, ranging from 21 and 22 points in British Columbia and Manitoba respectively to 41 points in Quebec and Newfoundland and Labrador.

Socioeconomic status is related to performance but to a lesser extent in Canada when compared to other countries

How equitable the benefits of schooling are to students from different socio-economic backgrounds can be understood from the relationship between socio-economic background and student performance. Socio-economic status (SES) is a term used to summarize a variety of factors, including parental education and occupation, which influence student performance. In PISA 2006, SES was measured by an index that includes information describing family structure, parental education and occupation, parental labour market participation and whether a student's family has specific educational and cultural possessions at home. This index is standardized to have an average of 0 and standard deviation of 1 across all OECD countries.

The averages of SES for Canada and the provinces are reported in Table 3.3. As was the case in previous PISA assessments, the average student in Canada had a relative socio-economic advantage compared to 15-year-olds in all OECD countries with a score on the mean index being higher than the OECD average of 0. Students in the four Atlantic provinces had scores below the Canadian average while the remaining provinces had scores that did not differ significantly from the Canadian average. Though the differences were not significant, Ontario, Alberta and British Columbia had higher scores than Canada. While students in all provinces tend to be more advantaged than 15-year-olds in other OECD countries, the differences across provinces suggest that all students in Canada may not have access to the same resources or opportunities.

The relationship between SES and performance can be measured by the average performance levels by quarter of the index. Table 3.3 shows the relationship between student performance and SES; students are grouped into 4 quarters based on their score on the SES index, those in the bottom quarter have lower levels of SES while those in the highest quarter have higher levels of SES.

As shown in Table 3.3 there is a positive relationship between SES and science performance. In Canada, those in the top quarter of SES score 68 points higher, equivalent to one proficiency level higher, than those in the lowest SES quarter. The effect of SES on performance is much lower than that observed across all OECD countries where the difference between those in the top and bottom quarter was 119 points.

At the provincial level, the SES performance gap between the bottom and top quarters ranged from 55 and 56 points in Prince Edward Island and British Columbia respectively to 93 points in Newfoundland and Labrador. The fact that all provinces had SES performance gaps below the OECD average and average performance above the OECD average in science can be an indication that jurisdictions are able to mitigate the effects of SES and achieve relative equity in the distribution of learning performance. However, as seen by provincial differences in performance related to SES, it appears that there is still room for improvement. Across jurisdictions, British Columbia represents an example of a jurisdiction that has minimized the relationship between student SES and performance while still maintaining high levels of student performance.

Summary

This chapter examined the relationship between various student background characteristics and achievement.

While no gender differences exist on the combined science scale, the performance patterns on two sub-scales suggest that boys and girls have very different levels of performance in different areas of science. It appears that boys may be better at mastering scientific knowledge whereas girls may be better at seeing the larger picture that enables them to identify scientific questions that arise from a given situation. Additionally, girls continue to outperform boys in reading while boys outperform girls in math.

Immigrant status, parental education and socio-economic background are all related to performance but to a lesser extent in Canada compared to OECD countries as a whole. However, differences observed across the jurisdictions, suggest that further improvements are possible. For example, Alberta represents a jurisdiction with higher performance and no differences in performance between immigrants and non-immigrant Canadians and British Columbia represents an example of a jurisdiction that has minimized the relationship between student SES and performance while still maintaining high levels of student performance.

While this chapter has only looked at the relationship between student background characteristics examined individually on performance, further multivariate research can determine the relative performance of student background characteristics impacting performance.

Note

1. Building on our Competencies: Canadian Results of the International Adult Literacy and Skills Survey, HRSDC and Statistics Canada 2005.