Linguistic Capital and Academic Achievement of Canadian- and Foreign- Born University Students

By Grayson, J Paul

Au Canada, plusieurs universites prennent des mesures pour recruter des immigrants ou leurs enfants et satisfaire leurs besoins- et parmi eux plusieurs ont l’anglais comme langue seconde. Il n’y a pas de recherches au Canada qui comparent la progression potentielle du capital linguistique des etudiants ayant l’anglais comme langue seconde et celui des autres etudiants au fil de leur parcours universitaire, avec les relations entre les progressions du capital linguistique et de l’acquisition des connaissances. L’auteur montre dans cette etude que, contrairement aux etudiants canadiens et ceux nes a l’etranger pour lesquels l’anglais est la premiere langue, le capital linguistique des etudiants nes a l’etranger dont l’anglais est une langue seconde s’accroit au cours des quatre annees d’etudes universitaires. Cependant, cette augmentation du capital linguistique ne correspond pas a une augmentation de l’acquisition des connaissances. In Canada, many universities are taking steps to recruit and meet the needs of immigrants and/or their sons and daughters, many of whom have English as a second language (ESL). There is, however, no research in Canada comparing potential increases in the linguistic capital of ESL and other students over the course of their university careers and the connection between increases in linguistic capital and academic achievement. In this study, it is shown that in contrast to Canadian- and foreign-born students for whom English is a first language, and Canadian-born ESL students, the linguistic capital of foreign-born ESL students increases over 4 years of university study; however, this increase in linguistic capital is not paralleled by an increase in academic achievement.

ALMOST 2 MILLION IMMIGRANTS ARRIVED IN Canada in the 1990s. Seventeen percent of this number were children aged 5-16. As a result of this migration pattern, the foreign born now make up 18.4 percent of the Canadian population (Canada 2003). Many newcomers have taken up residence in Canadian cities. In Toronto, Montreal, and Vancouver, Canada’s largest cities, immigrants comprise 44 percent, 18 percent, and 38, percent, respectively, of the population (Justus 2004). In earlier decades, most immigrants to Canada in general, and Toronto in particular, were of European descent. At the turn of the current century, however, most immigrants to Toronto were from China, India, Pakistan, the Philippines, Sri Lanka, Hong Kong, Iran, Russia, South Korea, and Jamaica (p. 43). In 2003, only 10 percent of immigrants had either English or French as a first language (Canada 2004).

Immigration patterns such as these result in well-documented economic (Samuel and Bassavarajappa 2006), social (Laguian, Laguian, and McGee 1997), and psychological (Beiser and Fang 2006) problems for immigrants. Moreover, the well-educated share in the difficulties. For example, national data indicate that in the 1990s, as many as one in four new immigrants with university degrees held jobs that required no more than secondary school education. This was twice the proportion of those born in Canada. In addition, the unemployment rate for new immigrants in the 25-54 age group was three times that of the Canadian born (Galaraneau and Morissette 2004).

Although large numbers of immigrants have postsecondary education (PSE), the results of the International Adult Literacy and Skills survey indicate that many well-educated immigrants have English language deficiencies. For example, whereas 37 percent of native- born university-educated Canadians scored in the highest literary category, the figure for similarly educated recent immigrants was only 11 percent. Even well-educated immigrants with English as a first language scored lower than the Canadian born. Importantly, length of residence in Canada was of no consequence for the literacy performance of immigrants. Evidence such as this suggests that any programs in place designed to facilitate literacy improvements among immigrants may be having less-than-optimal impacts. Understandably, immigrants with high levels of English literacy were most likely to be employed (HRDC 2005).

While it is difficult to say with certainty that they were inspired by findings such as the foregoing, it is nonetheless clear that procedures have been put in place in many Canadian postsecondary institutions designed to facilitate the adaptation of non-English-speaking new comers, or their sons and daughters, to Canadian society. For example, postsecondary educational institutions increasingly recognize that it is necessary to put in place processes that will ensure the recruitment of students from diverse backgrounds and the development of programs to meet the needs of such students. In Toronto, the city that hosts a plurality of Canada’s new immigrants, all three universities have policies designed to meet the needs of students from a diversity of backgrounds. As shown by Grayson (2007), the University of Toronto recognizes that, “our recruitment process [must] … be sensitive to the needs and interests of those whom we are attempting to recruit” (Neuman 2003). Potential recruits include those for whom English is a second language. York University has a similar commitment. As a result, its academic plan has, as one of its goals, “improving support for students in need of additional support, and students for whom English and French are second languages” (Anonymous 2005). Because of its similar commitment to diversity, the academic plan at Ryerson University recognizes that, “there may be many implications for program design and delivery, particularly related to factors such as learning styles and English language familiarity” (Anonymous 2003).

How well are such policies working? Although some research shows that immigrants are less likely than others to find employment equal to their level of education, and that students for whom English is a second language do not graduate from high school in numbers comparable to those of English speakers (Radwanski 1987; Derwing et al. 1999; Roessingh and Watt 2001), a national study indicates that after imposing controls for variables such as income, gender, visible minority, and immigrant statuses, those with a mother tongue other than English or French are 12 percent more likely to be enrolled in universities (Frenette 2005). Mother tongue is of no consequence for college registration.

While these findings may appear to be discrepant, they can be interpreted as follows: As a group, highly educated English as a second language (ESL) immigrants are less likely than others to find employment consistent with their education. In addition, ESL students, whether immigrants themselves or sons and daughters of immigrants, may have lower high school graduation rates than others. This said, ESL students are a diverse group. Members of some groups, such as those of Chinese origin, may be inclined to graduate from high school, and proceed to higher education, in great numbers. Indeed, the proportion of such individuals who proceed in this fashion may be sufficiently high to result in the finding that ESL students in general are more likely than others to enroll in universities. This interpretation is consistent with analyses of the educational levels of members of different groups of young people based on census materials. For example, in Toronto, within the 25- 34 age group, 33.2 percent and 28.3 percent of European origin males and females, respectively, have a university degree. The comparable figures for those of Chinese origin are 48.5 percent and 54.3 percent (Ornstein 2006, table 3.a).

Although they may enter PSE in greater-than-average numbers, it does not mean that postsecondary experiences of ESL immigrants or their sons and daughters are similar to those of other groups. For example, it was found in one study of a Canadian university that ESL students with 3-5 years of experience in Canadian high schools performed worse academically than either non-ESL students or ESL students who were required to pass an English language proficiency test before being admitted. The reason for the relative success of the latter group is in part attributed to the fact that they were eligible for English language support programs. By comparison, such programs were not available to ESL students coming from Canadian high schools (Fox 2005).

Other research indicates that when they enter postsecondary institutions, the dropout rates for ESL students may be no higher than for other students; however, consistent with Fox (2005), the grades of the former are relatively low (Roessingh et al. 2004). Importantly, in a study in which the grades of ESL students were compared with those of students who spoke English while growing up, it was also found that after controls had been applied for high school grades, in addition to home language, university grades were also affected by ethno-racial origin (Grayson 2008). Additional research on international and domestic ESL and English-speaking students at the University of British Columbia, York University, McGiIl University, and Dalhousie University found that international ESL, domestic ESL, international English speakers, and domestic English speakers, in that order, experienced difficulties communicating in English. Again, after controls had been applied for high school grades, the first-year grades of students for whom English was a second language were lower than those of students with English as a first language (Grayson and Stowe 2005). Although there is mounting evidence that independent of high school grades ESL students, both domestic and international, face English problems at the postsecondary level that are concomitant with relatively low levels of achievement, little is known about the extent to which the university experience contributes to the development of linguistic capital of ESL and non-ESL students from different social backgrounds, and the connection of linguistic capital to academic achievement. This article will fill in some of these lacunae.

ORIENTATION

The examination of the issues noted above can be viewed in the more general context of cultural reproduction (Grayson 2008). For Bourdieu and Passeron (1990), “linguistic capital” is a component of “cultural capital” that predisposes the sons and daughters of the privileged classes to academic success; however, a clear definition of the concept is not provided. Other writings of Bourdieu do not deal with this limitation. Using Bourdieu’s concepts as a point of departure, however, Sullivan (2001:893) defines linguistic capital as, “the ability to understand and use ‘educated’ language.” Morrison and Lui (2000:473), in an examination of colonial Hong Kong, define linguistic capital as, “fluency in, and comfort with, a high-status world-wide language which is used by groups who possess economic, social, cultural and political power and status in local and global society.” As in the former colony of Hong Kong, in English Canada, the dominant language is English.

Despite Bourdieu’s inattention to a precise definition of linguistic capital, he makes four things very clear. First, linguistic capital, as a component of a broader cultural capital, is acquired in large part from parents. Second, the uneducated lower classes in capitalist societies lack the kind of linguistic capital valued by the powerful. Third, academic success is contingent upon the acquisition of linguistic capital. Fourth, deficiencies in linguistic capital contribute to the winnowing out of the sons and daughters of the less advantaged from successive levels of education. This means that relatively few students from disadvantaged backgrounds are found at the university level.

Bourdieu and Passeron argue that in elementary schools teachers constantly evaluate students’ use and understanding of language. These evaluations find lower-class students deficient. Not only teachers’ evaluations but also tools of assessment themselves (e.g., IQ and other standardized tests) have been found to embody ways of viewing the world foreign to economically disadvantaged students. Because of such biases, the performance of linguistically disadvantaged students on many tests is such that the educational credentials necessary for success in capitalist societies elude them (Gipps 1999).

At postsecondary levels of education, Bourdieu and Passeron argue that it is difficult for students with relatively limited linguistic capital to understand the “convoluted” language of their professors (e.g., the use of “big” words and compound-complex sentences). This difficulty would be compounded for ESL students in English Canada. Overall, the effect of processes such as these is that “the educational mortality rate … increase[s] as one moves towards the classes most distant from scholarly language” (Bourdieu and Passeron 1990:73).

Bourdieu’s argument that linguistic capital is a function of class, and is used to marginalize the lower classes at all levels of education, is not without its critics. It has been argued, for example, that in France language and high culture are more a vehicle of social distinction than in other countries such as Britain and the United States (Dumais 2002). Consequently, a lack of linguistic capital may be of more consequence in France than elsewhere. Also, it may be naive to assume that in Britain, teachers, many of whom have lower-class roots, are gatekeepers for the upper classes (Sullivan 2001; Dumais 2002; Robson 2003). Despite these and other criticisms, it is clear that language development varies by class and that disadvantage accrues to the relatively underprivileged (Ensminger and Slusarcik 1992; Bayder, Brooks-Gunn, and Furstenberg 1993; Parcel and Menaghan 1994; Hart and Risley 1995; Brooks-Gunn, Klebanov, and Duncan 1996; Alexander, Entwisle, and Horsey 1997; Farkas and Beron 2001).

In Canada, there is some research indicating that the cultural capital (which includes its constituent linguistic capital) available to students has implications for progression to higher education and for income upon graduation from university. As yet, however, the extent to which cultural capital affects postsecondary performance has received little attention.

As an example of the first point, research conducted in Canada has shown that the cultural capital of parents is of consequence for the completion of postsecondary studies by their children. For example, a longitudinal examination conducted in British Columbia revealed that 5 and 10 years after high school graduation individuals whose parents could be seen as being rich in cultural capital were far more likely than others to have completed postsecondary studies (Andres and Grayson 2003). It is worth noting that in this study 5 years after high school graduation the effect of the cultural capital of parents on the occupational status of children took an indirect route via the educational attainment of their children. There was no direct connection between parents’ cultural capital and the occupational status of their children. An example of the second point is provided by research in which the impact of cultural capital, human capital, and social capital on incomes 2 years after graduation was analyzed. The results of the study indicate that the cultural capital of parents was passed on to university graduates in the form of income levels after graduation (Grayson 2004). In short, all else being equal, the incomes of graduates from families high in cultural capital were higher than those of students from low cultural capital families.

Within the university, a manifestation of the operation of cultural and linguistic capital has been demonstrated in a Canadian study focusing on the grades of different language groups after adjustments had been made for fathers’ educational background, age, ethno-racial origin, high school grades (from administrative records), self-assessed communication skills, and faculty of enrollment. By and large, the university grades (from administrative records) of the sons and daughters of the Canadian born with English as a first language were higher than those of the children of immigrants, most of whom spoke a language other than English while growing up (Grayson 2008).

While the foregoing study indicated a disadvantage on the part of nonEnglish linguistic groups, it could not examine possible changes within and between groups of different students once they enter university. More concretely, does the linguistic capital of various linguistic groups increase over the course of university studies? University professors might hope so. Also, does the academic performance of initially disadvantaged linguistic groups improve over time? As indicated earlier, answers to these questions are the focus of the current inquiry.

SAMPLE

York University, the institution on which the current study is based, is a racially diverse comprehensive commuter university of approximately 45,000 students located in Toronto, Canada. At the end of their first year of study in 1995, a mail-out questionnaire was returned by a random sample of 1,865 students who had entered York University directly from high school in 1994. These 1,865 students represented a response rate of 64 percent. Exactly the same questionnaire was mailed to this original group of respondents at the end of 1996,1997, and 1998. By the final year, 513 students (or 28 percent of the original sample) had responded to each wave of the study. When adjustments are made for the fact that in the intervening years students had left the university either before or after degree completion, the 513 students who completed the final questionnaire represent 55 percent of those who had responded to the original survey and who were still enrolled in the university and eligible to participate (Grayson 1999). In a longitudinal study of this nature, this is an excellent retention rate (Dey 1997).

Given that in all faculties with the exception of Arts for each wave of the study questionnaires were sent to all students who originally enrolled in the faculty in 1994, it was not possible to compare all ongoing survey participants with a broader base of students. In view of the large size of Arts, however, it was possible to compare Arts students who had remained in the study to a sample of Arts students in general. When this was done, it was found that ongoing participants were similar to Arts students in general in terms of ethno-racial origin, sex, family income, number of completed credits, and grade point average (GPA) (Grayson 1999:48).

MEASURES

In view of the potential importance of English linguistic capital to academic success, survey respondents were asked to self-assess a number of different English communication skills. As noted by Portes and Schauffler (1994), unlike self-assessments of other skills, those of language ability have been found to be both reliable and valid. Consistent with this conclusion, an examination of data derived from The International Adult Literacy Survey (IALS) (Anonymous 1995) jointly undertaken by Statistics Canada and The Educational Testing Service sheds important light on the relationship between self-reported and other measures of skills. As noted elsewhere (Grayson 2008), this study obtained self- assessments of reading, writing, and quantitative skills, and required participants to complete a number of performance tasks of varying levels of difficulty related to prose literacy, document literacy, and quantitative literacy. For Canadians aged 16-25 who had completed PSE (the group most comparable to the subjects of the current study), the correlations between self-assessed reading and testmeasured prose and document literacy were .45 and .36, respectively (analysis of data by author). The correlations between self-reported writing skill and test-measured prose and document literacy were .29 and .22; however, the last mentioned was not statistically significant. Finally, there was a correlation of .29 between self-assessed quantitative skills and quantitative literacy as measured in tests. This study, based on Canadian data, shows that there are small-tomoderate statistically significant correlations between self-assessments of communications skills and skills measured in other ways. While the magnitude of these correlations is insufficient for making academic decisions about, for example, the linguistic capital of individual university students, they suggest that in comparisons between groups self-reports may be used as a cautious proxy for more “objective” measures. This assumption is supported by the results of a great deal of other research (Berdie 1971; McMorris and Ambrosino 1973; Pohlmann and Beggs 1974; Baird 1976; Dumont and Troelstrup 1980; Pascarella and Terenzini 1991; Evers et al. 1993; Pike 1994, 1995a, 1995b; Kuh et al. 1997). The National Survey of Student Engagement, which is now used widely to measure experiences and outcomes in Canadian universities, relies exclusively on self-assessments of linguistic and other skills.1

In the current study, English linguistic capital was assessed by four questions in which students were asked whether they agreed or disagreed that it was easy for them to engage in different English language activities. The areas identified were: speaking English; reading English; writing in English; and following a conversation in English. Response options ranged from 1, meaning that it was very hard to complete the activity in the statement, to 5, indicating that it was very easy.

Cronbach’s a2 for the four questions combined was .71, .71, .77, and .71 for each of the 4 years. By removing the results of the question pertaining to ease of writing in English, however, alpha increased to .87, .87, .90, and .89. As a result, although the alpha for all four items was sufficient, it was decided to combine speaking, reading, and conversing into one index (nonwritten skills) and treat the item dealing with ease of writing in English (written skills) separately from the other three items. The average nonwritten score in the first year was 4.7 with a standard deviation of .59. The respective figures for written scores were 4.3 and 1.3.

Given the fact that it has been shown in many studies that the cultural and linguistic capital of children is directly related to the class and education of their parents (Laureau and Weininger 2003; Robbins 2005), students were asked to provide information on parental education. For the current study, students were classified in terms of having a minimum of one parent who had at least some postsecondary experience. Overall, 57 percent of students reported that at least one parent had at least some postsecondary experience.

Country of birth was determined by a question in which students were simply asked to identify the country in which they were born. In total, 75 percent of students were born in Canada. For students born overseas, the average number of years in Canada was 3.0 with a standard deviation of 5.9. These figures indicate that a large portion of foreign-born students were relatively recent immigrants.

Language background was assessed by asking students to identify the language they spoke in the home while growing up and still understood. Sixty-seven percent of students reported English as their first language. In the analyses, those who specified English were categorized as Englishspeaking students. Students who identified a language other than English were categorized as ESL.

Of the 513 students who completed all four surveys, 59 percent were born in Canada and spoke English while growing up. Eight percent of respondents were born outside of Canada and also spoke English while growing up. Large numbers of these were immigrants from Caribbean countries. A further 16 percent were born in Canada but spoke a language other than English in their childhood homes. The remaining 17 percent were born outside of Canada and spoke a language other than English while growing up. With the exception of Canadian-born English speakers (303 cases), the absolute number of cases in some groups is relatively low. Overall, 41 English speakers were born outside of Canada. Eighty-seven ESL students were born outside of Canada and 82 ESL students were born in Canada. As a consequence of these relatively small group sizes, for some statistical analyses, false-negative findings are possible.3 Discussions of results will take this possibility into consideration.

Information on high school grades (Ontario Academic Credits [OAC]), university GPA, and faculty of enrollment was obtained from administrative records and linked to survey data. The average high school grade was 80.3 percent with a standard deviation of 6.3 percent. In the first year, the average GPA was 5.9 (low of O and high of 9) with a standard deviation of 1.5.

At the time of the study, York had several faculties, the largest of which was the Faculty of Arts. Medium-sized faculties included Science and Engineering, Fine Arts, and Glendon College (a small bilingual college located 19km away from the main campus). Environmental Studies and the Schulich School of Business were relatively small faculties, at least at the undergraduate level.

ANALYSIS

Analysis will proceed in three steps. First, as other studies have shown that grades in Science tend to be lower than in other faculties (Astin 1993; Pascarella and Terenzini 2005), we will focus on the enrollment patterns of different language groups. This is a necessary step as otherwise we might attribute low grades to particular language groups when in fact such grades reflect the fact that such language groups enroll in particular faculties in large numbers. second, through repeated analysis of variance, a method that allows us to assess changes between individuals and groups over time, potential changes in English linguistic capital and academic achievement will be examined over the 4 years of the study. Third, regression analyses will be carried out to determine, in each year, the effects of variables described above on academic achievement (GPA).

FACULTY DISTRIBUTIONS

The distribution of language groups across faculties can be found in Table 1. For current purposes, it is sufficient to make four observations. First, the largest percentage of English Canadian students (34 percent) are enrolled in the Faculty of Arts. The second largest group of these students (21 percent) is enrolled in Fine Arts.

A large percentage of Canadian ESL students (43 percent) is also enrolled in the Faculty of Arts. By comparison, the largest single group of foreign English students (30 percent) is enrolled in Science and Engineering. This faculty is also home to the largest group of foreign ESL students (36 percent). Given the bilingual nature of Glendon College, it is not surprising that a substantial number of Canadian ESL students (20 percent) are found there. While differences among groups are not always large, they are statistically significant and indicate that different language groups are not randomly distributed across all faculties.

Repeated Analyses of Variance

Because of some small cell sizes noted earlier, it is not practical in the repeated analyses of variance to control for a wide variety of variables. (This task will be left for the regression analyses.) Despite this limitation, repeated analyses of variance allow us to examine the possibility of change in nonwritten and written linguistic capital and academic achievement over 4 years of study.

Table 1

Faculty of Enrollment by Language Group

Figure 1. Nonwritten English Linguistic Capital Score by Birthplace and First Language

The average scores (maximum of 5) for nonwritten and written linguistic capital of each language group are plotted in Figures 1 and 2. The first point of interest in Figure 1 is that with the exception of students born outside Canada who did not speak English while growing up (foreign ESL), for each group, the absolute scores for nonwritten English linguistic capital are reasonably high. For example, in year 1, students who were born in Canada and who spoke English while growing up (Canadian English) scored 4.91; students who were born overseas but spoke English in their families (foreign English) self-rated at 4.79. For Canadian-born students who spoke a language other than English while growing up (Canadian ESL), the score is 4.80. The score for students born overseas who did not speak English in their households (foreign ESL) is 4.10.

The second point of interest is that with the partial exception of foreign ESL students, the scores of different language groups remain basically the same over the 4 years of study. For example, Canadian English students score 4.91 in year 1 and 4.94 in year 4. By comparison, the year 1 score for foreign ESL students is 4.10 and modestly increases to 4.25 in year 4.

As might be expected, given the magnitude of change over 4 years, and the relatively small numbers in some groups, the results of repeated analyses of variance on these and other variables reported in Table 2 indicate that, overall, wii/im-subject variance (changes within individuals from one survey to the next) over the 4 years of the study is not statistically significant, nor are changes for individual groups of students. Information in Table 2 also shows that over the 4-year time span, the between-subject differences (differences between the groups averaged over the entire 4 years) among groups are statistically significant in nonwritten English scores. In addition, the scores for Canadian ESL and foreign ESL students are statistically significant from those of Canadian English students, the reference group. By comparison, and as might be expected, differences between the scores of the reference group and foreign English speakers are not statistically significant. Recall, however, that the absolute numbers in the latter group are small, and false negatives are possible. Figure 2. Written English Linguistic Capital Score by Birthplace and First Language

The results for English linguistic capital as measured by the single question are summarized in Figure 2. On this dimension, scores for all groups are lower than for nonwritten English. Most noticeable differences for the first 3 years are between foreign ESL students and all other groups. For example, in year 1, the respective scores for Canadian English, foreign English, Canadian ESL, and foreign ESL students are 4.52, 4.00, 4.26, and 3.61, respectively. In year 4, there is some convergence. As was the case in the examination of nonwritten English, the greatest gains in written English capital are made by foreign ESL students who increase from a score of 3.61 in year 1 to 4.00 in year 4.

Table 2

Significance of Repeated Analyses of Variance

The results of the repeated analyses of variance reported in Table 2 indicate that the overall changes from 1 year to the next (within-subject changes) for written English linguistic capital are not statistically significant nor are the gains made by Canadian and foreign speakers of English and Canadian ESL students. By contrast, improvements in the written English of foreign ESL students are statistically significant. Put differently, there is an improvement in this aspect of the linguistic capital of foreign ESL students over the study period. In addition, the overall differences among groups averaged over 4 years (between-subject differences) are statistically significant. Also, differences between Canadian English students and those in each of the other three groups are statistically significant.

Keeping in mind the potential limitations imposed by small numbers in some groups, taken together, these two sets of findings indicate that over time there are no changes in the nonwritten English linguistic capital of students. By comparison, the written English capital of foreign ESL students does improve. In addition, with the exception of nonwritten English linguistic capital scores of foreign English students (which again may be a reflection of low numbers), between-subject differences between Canadian English and other groups in nonwritten and written English averaged over 4 years are statistically significant.

Figures. Grade Point Average (GPA) by Birthplace and First Language (Adjusted for Ontario Academic Credits)

As in Canada the best predictor of academic achievement in first- year university is high school grades (Grayson 1996), it is important in analyses of university achievement to control for high school achievement. If this step is not taken, we may wrongly attribute university achievement to ascriptive factors, such as language, when in fact achievement is not a product of language but of the fact that different language groups enter university with different levels of previous achievement.

In the current study, Canadian English and foreign ESL students enter university with comparable high school averages (roughly 81 percent). By comparison, the high school grades of foreign English and Canadian ESL students are slightly lower. Differences, however, are not statistically significant: high school grades do not vary by language group.

The annual grades achieved by different groups of students with adjustments for high school grades (OAC or the equivalent) are summarized in Figure 3. As can be seen, there are considerable differences in the grades of different language groups. In year 1, the highest grades, 6.04 (the scheme goes from O to 9 at York), are achieved by Canadian English students, and the lowest, 4.35, by Canadian ESL students. The GPAs of foreign English and foreign ESL students are 5.43 and 5.66, respectively.

The within-subject analyses of variance reported in Table 2 indicate that overall increases in grades between years 1 and 4 from one year to the next are statistically significant; however, when groups are examined individually, only changes for Canadian English students are statistically significant. Their grades go from 6.04 to 6.41. Once again, however, the absence of a statistically significant change in other groups must be seen through the lens of small numbers for some groups. We can see from Table 2 that differences among groups averaged over 4 years are statistically significant; however, while differences between the reference group and foreign ESL students are statistically significant, other comparisons are not.

It was noted earlier that in analyses of variance, the small size of some groups may have resulted in some false-negative findings. As a result, a conservative approach will be adopted for the overall interpretation of results discussed in this section. More specifically, it will be assumed that statistically significant results reflect real differences in the population while results that are not statistically significant, because of some small group sizes, will be viewed as inconclusive. Bearing in mind these caveats, there are four overall conclusions that can be derived from the analyses of variance. First, averaged over the entire study period, Canadian-born speakers of English enjoyed a higher level of nonwritten linguistic capital than Canadian- and foreign-born ESL students. second, averaged over the entire 4 years, Canadian English students displayed an advantage over all other groups in written linguistic capital. Nonetheless, foreign ESL students experienced an increase in written linguistic capital from one year to the next. Third, averaged over the total study period, the academic achievement of foreign ESL students was lower than for the Canadian English group. Fourth, Canadian-born English-speaking students experienced an increase in academic achievement from one year to the next.

Regressions

The foregoing repeated analyses of variance are appropriate in examining change in linguistic capital and achievement over time; however, they do not allow us to examine the degree to which cultural capital and language group have independent effects on academic achievement, controlling for other theoretically relevant variables. This objective is realized through regression analyses. In these, sample size does not present the same potential problems as it did for analyses of variance: with 12 independent variables, the sample size of 513 meets the minimal requirements for this type of analysis (Brace, Kemp, and Snelgan 2003).

The results of the regression analyses are given in Table 3. Note that “foreign English,”"Canadian ESL,” and “foreign ESL” are dummy variables, with 1 indicating group membership and 0 indicating nonmembership. Canadian-born English-speaking students are the reference group. Dummy variables have also been created for various faculties at York with Arts as the reference category. For these variables, 1 was recorded for enrollment in the faculty and a value of 0 was given for nonenrollment. All of the other variables have been explained in a previous section. In the analysis, variables were entered in blocks in order to identify changes in explained variance attributable to different groups of variables. The cumulative amount of explained variance is given after the introduction of parental education, OAC grades, language group variables, faculty variables, and for nonwritten and written English linguistic capital.

Table 3

Regression Coefficients (beta) for GPA

In analyzing data, it is best to consider beta weights for the same variable across all four surveys. When doing this, it is evident that beta’s associated with having at least one parent with a PSE (a likely indication of a relatively high cultural capital) are small, have positive signs in some years and negative in others, and are not statistically significant. Not surprisingly, this variable explains a negligible amount of variance in GPA. In essence, the indicator of cultural capital used here reveals no independent effect of this variable on academic achievement in any year.

When the effect of OAC (high school) grades is examined, as expected, it has a relatively large, positive, and statistically significant impact on grades in each year. Indeed, the impact of OAC grades on GPA is the largest of any variable. In year 1, OAC grades account for 29 percent of the explained variance (R^sup 2^). The effect of high school achievement is still felt in year 4 in which it explains 16 percent of the variance.

As was anticipated, the signs of the beta’s for the variable foreign English are negative, indicating that the GPAs of this group of students are slightly lower than those of the reference group. This said, for no years are the coefficients statistically significant, and for all years they are low.

In every year, with the exception of year 4, the coefficients for Canadian ESL status are negative and small. This finding suggests a slight disadvantage to this group of students compared with Canadian English students; however, none of the beta’s is statistically significant. Similarly, as anticipated, all of the coefficients for foreign ESL status are negative. Once again, however, with the exception of year 3, the coefficients are not statistically significant. For all years they are low. Collectively, the amount of variance attributable to language group ranges from a high of approximately 2 percent ((.307-.291) x 100) in year 1 to a low of .2 percent in year 4. In contrast to the negligible effect of language group on GPA, many of the effects of faculty of enrollment on grades are statistically significant. In year 1, students enrolled in Environmental Studies (beta = .084) received slightly higher grades than those in Arts. Also, in the Schulich School of Business, in year 1 (beta = .092), grades were higher than those in Arts. Most evident, however, is that students in Science (beta = – .219) received grades that were lower than those in Arts. Students enrolled in Glendon received slightly higher grades (beta = .082) than their counterparts in Arts in year 1.

In year 2, the grades of Fine Arts Students exceeded those of their peers in Arts (beta = .128), while once again the grades of Science students were relatively low (beta = – .128). In year 3, students in all faculties received grades similar to those of Arts students. In year 4, however, once again, the grades of Science students were relatively low (beta = – .120).

The amount of variance in grades explained by faculty of enrollment ranges from a high of 8 percent in year 1 to a low of 2 percent in year 4. Overall, faculty of enrollment explains far more of the variance in grades than language group.

For nonwritten linguistic capital, only coefficients for years 3 (beta = .096) and 4 (beta = .195) are statistically significant. These figures indicate a small positive impact of nonwritten linguistic capital on achievement in these years. While in year 4 nonwritten linguistic capital increases total explained variance by 2 percent, the increase in variance in other years approaches zero.

An examination of the final variable in the regressions, written linguistic capital, indicates small negative coefficients for all years; however, none is statistically significant, indicating that high levels of written linguistic capital do not translate into high levels of achievement.

The table R^sup 2^’s indicate that the total explained variance for the regression models ranges from a high of 39 percent in year 1 to a low of 22 percent in year 4. R^sup 2^’s of these magnitudes are typical for analyses of academic achievement in which controls for high school grades are imposed (Pascarella and Terenzini 2005).

The overall impression left by the data in Table 3 most relevant to the current undertaking is as follows: As seen earlier, while family cultural capital in Canada may predispose students to higher education, and has implications for earnings after graduation, it is of no discernable consequence for academic achievement, at least at this university. This finding may be explained by the possibility that independent of parental education, all families of students had a level of cultural capital sufficient to inspire their children to enroll in university.

Second, while, as expected, belonging to a group other than Canadian English results in depressed levels of academic achievement, the effect is confirmed only for foreign students in year 3, in which this group earns slightly lower grades than the reference group.

Third, in contrast to language group, faculty of enrollment has considerable consequences for academic achievement. Most noticeable is that students enrolled in Science receive lower grades than those in Arts for 3 of the 4 years. Where statistically significant, the data show that students in other faculties receive higher grades than their peers in Arts.

While nonwritten linguistic capital, as expected, results in slightly higher grades in years 3 and 4, written linguistic capital is of no consequence for grades. There are a number of possible explanations for this finding.

First, despite the fact that, as shown in the repeated analyses of variance, all language groups have lower written linguistic capital than Canadian English students, their levels may be sufficient to meet the academic challenges of university life. Moreover, excluding exams, students have time to think about, and modify, their written work before submission. As a result, deficiencies in written linguistic capital may have relatively muted effects on achievement.

A second explanation involves the possibility that in their attempts to accommodate, professors, to a degree, overlook deficiencies in the written work of students for whom they know English is a second language. This explanation is consistent with research at the secondary school level in which it was found that teachers frequently award what are called “goodwill” grades to ESL students, in the belief that in doing so they are providing needed support (Roessingh and Kover 2003).

A third possible explanation is that students who have written linguistic difficulties stay away from courses in which they would be required to write essays and reports. Instead, they might seek out courses in which considerable weight is given to performance on multiple-choice tests.

A fourth possibility is that the measure of written linguistic capital used in this study underestimates the real capital possessed particularly by ESL students. At this point, further research is needed to determine which of the above possible explanations is most probable.

CONCLUSIONS

Most recent immigrants to Canada have limited English linguistic capital. While such deficiencies are associated with un- and underemployment of immigrants with higher education, a disproportionate number of immigrants, or their sons and daughters, proceed to university. We know from previous research that such students receive relatively low grades once they enter university; however, we do not know whether the university experience results in an increase in their linguistic capital, or the complete nature of the relationship between linguistic capital and academic achievement.

The results of the current research suggest several conclusions with regard to these matters. First, over 4 years of study, by and large, Canadian English students maintain an advantage in linguistic capital over other groups of students. second, bearing in mind possible limitations imposed by some small numbers, with the exception of written linguistic capital for Foreign ESL students, there is in general no increase in the linguistic capital of students over 4 years of study. In other words, despite the probability of an increase in subject matter expertise and other forms of human capital over the course of a university education, there is no improvement in linguistic capital. Third, nonwritten linguistic capital is of little consequence for academic achievement in university. Written linguistic capital is also of no consequence for academic achievement. When considering the implications of this finding, however, it is important to remember that despite differences among different language groups, students may nonetheless enter university with sufficient linguistic capital to meet academic demands. If they get low grades, it is because of other factors. Alternately, students with low levels of linguistic capital may manage their course selections in ways that minimize any linguistic deficiencies that they might have. Fourth, while family cultural capital has been shown to affect whether or not young adults attend postsecondary institutions, and is of consequence for earnings of university graduates, it is of no consequence for achievement in university. This finding may be explained by the possibility that students acquire sufficient cultural capital to meet the challenges of university life in ways other than through parental transmission. Fourth, in contrast to cultural capital and linguistic capital, academic achievement in high school is the best predictor of achievement in university. This finding may indicate that along with subject matter expertise, in high school students acquire sufficient cultural capital to offset any potentially negative effects of capital deficits in the family. Fifth, the second best predictor of academic achievement is faculty of enrollment. In other words, all else being equal, students who enroll in, for example, Science, receive relatively low grades.

As seen in the introduction, universities are taking steps to recruit minority and ESL students and to meet their needs. On the basis of the findings of this study it is difficult to be conclusive as to the results of their efforts, at least in the university under study. On the one hand, it seems that, with one exception (written English of foreign ESL students), there is no increase in the linguistic capital of students over the course of their studies. From an educational point of view, this is regrettable, and steps should be taken to remedy the situation. Moreover, while there are anecdotal explanations for this state of affairs, there appears to have been no systematic inquiry into the matter in Canada. On the other hand, it appears that membership in particular language groups does not affect students’ academic achievement. The grades that they get are more related to their level of ability and to the faculty in which they enroll. Unfortunately, this lack of connection between language group membership and linguistic capital on grades may indicate that within the university certain linguistic deficiencies are overlooked. While further research is needed to confirm this possibility, it is supported by anecdotal evidence.

LIMITATIONS

There are several limitations of the current study. First, the number of individuals in some linguistic groups is relatively small. As a result, there is a possibility of false-negative findings in some of the analyses of variances. Additional research with larger samples is needed to confirm the findings of the current study. Second, although many other studies have used education of parents as an indicator of cultural capital, education, and cultural capital are not synonymous. It is possible for individuals with low levels of education to have relatively high levels of cultural capital. The reverse is also true.

Third, self-assessments of linguistic capital have been used to good effect in numerous other studies, and, with some qualifications, studies using self-reports yield the same general results as those using “objective” measures. Nonetheless, studies utilizing objective as well as subjective measures of linguistic capital are desirable, albeit very expensive. As a result, Canadian studies utilizing more objective measures of cultural capital, and objective measures of linguistic capital, are needed.

Fourth, the current study was carried out in a university in which relatively large numbers of students have English as a second language and most come from families with relatively modest incomes. Also, large numbers are the first in their families to attend university. Under circumstances such as these, it might be that a cultural context develops that is more supportive of the needs of ESL and first-generation students than would be the case in universities in which such students were a small minority. Under circumstances such as these, the experiences and outcomes of different language groups may be different from that discovered in this study. As a result, research is needed to determine whether the results of the current study can be generalized to other circumstances. It should be noted that unless they involve very large samples so that individual institutions can be compared, national studies cannot achieve this objective.

1 This is not an endorsement of the use of the survey in Canada. While it is a valid and useful survey, its results are being used in many unjustifiable ways.

2 Cronbach’e alpha is a measure of the internal consistency of a scale, based on correlations among different items. Generally, an alpha of .6 to . 7 indicates an acceptable level of consistency.

3 Assuming a similar alpha and effect size, a targe sample is more likely to result in a statistically significant finding than a small sample.

REFERENCES

Alexander, K., D. Entwisle, and C. Horsey. 1997. “From First Grade Forward: Early Foundations of High School Dropout.” Sociology of Education 70(April):87-107.

Andres, L. and J.P. Grayson. 2003. “Parents, Educational Attainment, Jobs and Satisfaction: What’s the Connection? A 10-Year Portrait of Canadian Young Women and Men.” Journal of Youth Studies 6:181-202.

Anonymous. 1995. Literacy, Economy and Society: Results of the First International Adult Literacy Survey. Ottawa: OECD and Statistics Canada.

Anonymous. 2003. Learning Together: An Academic Plan for Ryerson University. Toronto: Ryerson University.

Anonymous. 2005. The Senate of York University Academic Plan: Academic Priorities 2005-2010. Toronto: York University.

Astin, A. 1993. What Matters in College? San Francisco, CA: Jossey Bass.

Baird, L. 1976. Using Self-Reports to Predict Student Behavior. New York: College Entrance Examination Board.

Bayder, N., J. Brooks-Gunn, and F. Furstenberg. 1993. “Early Warning Signs of Functional Illiteracy: Predictors in Childhood and Adolescence.” Child Development 64:815-29.

Beiser, M. and H. Fang. 2006. “Ethnic Identity, Resettlement and Stress and Depressive Effect among Southeast Asian Refugees in Canada.” Social Science and Medicine 63:137-50.

Berdie, R. 1971. “Self-Claimed and Tested Knowledge.” Educational and Psychological Measurement 31:629-36.

Bourdieu, P. and J.C. Passeron. 1990. Reproduction in Education, Society and Culture. London: Sage.

Brace, N., R. Kemp, and R. Snelgan. 2003. SPSS for Psychologists. Mahwah, NJ: Lawrence Erlbaum.

Brooks-Gunn, J., P. Klebanov, and G. Duncan. 1996. “Ethnic Differences in Children’s Intelligence Test Scores: Role of Economic Deprivation, Home Environment, and Maternal Characteristics.” Child Development 67:396-408.

Canada. 2003. Canada’s Ethnocultural Portrait: The Changing Mosaic. Ottawa: Industry Canada.

Canada. 2004. Facts and Figures 2004. Ottawa: Strategic Research and Evaluation, Citizenship and Immigration Canada.

Derwing, T.M., E. De Corby, J. Ichikawa, and K. Jamieson. 1999. “Some Factors that Affect the Success of ESL High School Students.” Canadian Modern Language Review 55:532-47.

Dey, E.L. 1997. “Working with Low Survey Response Rates: The Efficacy of Weighting Adjustments.” Research in Higher Education 38:215-27.

Dumais, S. 2002. “Cultural Capital, Gender, and School Success: The Role of Habitus.” Sociology of Education 75(January):44-68.

Dumont, R. and R. Troelstrup. 1980. “Exploring Relationships between Objective and Subjective Measures of Instructional Outcomes.” Research in Higher Education 12:37-51.

Ensminger, M. and A. Slusarcik. 1992. “Paths to High School Graduation or Dropout: A Longitudinal Study of a First-Grade Cohort.” Sociology of Education 65(April):95-113.

Evers, F., J. Rush, J. Krmpotic, and J. Duncan-Robinson. 1993. Making the Match: Phase 2. Guelph: University of Guelph.

Farkas, G. and K. Beron. 2001. “Family Linguistic Culture and Social Reproduction: Verbal Skill from Parent to Child in the Preschool and School Years.” Paper presented at the annual meeting of the Population Association of America, Washington, DC.

Fox, J. 2005. “Rethinking Second Language Admission Requirements: Problems with Language-Residency Criteria and the Need for Language Assessment and Support.” Language Assessment Quarterly 2:85-115.

Frenette, M. 2005. Is Post-Secondary Access More Equitable in Canada or the United States? Ottawa: Statistics Canada.

Galaraneau, D. and R. Morissette. 2004. Immigrants: Settling for Less? Ottawa: Statistics Canada.

Gipps, C. 1999. “Socio-Cultural Aspects of Assessment.” Review of Research in Education 24:355-92.

Grayson, J.P. 1996. Under- and Over-Achievement in First Year. Toronto: Institute for Social Research, York University.

Grayson, J.P. 1999. The Student Experience at York University: The Effects of Income, Race, and Gender over Four Years. Toronto: Institute for Social Research, York University.

Grayson, J.P. 2004. “Social Dynamics, University Experiences, and Graduates’ Job Outcomes.” British Journal of Sociology of Education 25:609-28.

Grayson, J.P. 2007. “Unequal Treatment and Program Satisfaction among Students of European and Chinese Origin.” Canadian Journal of Higher Education 37:51-85.

Grayson, J.P. 2008. “Language Background, Ethno-Racial Origin, and Academic Achievement of Students at a Canadian University.” International Migration 46.

Grayson, J.P. and S. Stowe. 2005. “Language Problems of International and Domestic ESL Students at UBC, York, McGill and Dalhousie, and Academic Achievement.” Paper presented at the annual meetings of the Canadian Society for the Study of Higher Education, London, ON.

Hart, B. and T. Risley. 1995. Meaningful Differences in Everyday Experience of Young American Children. Baltimore, MD: Paul Brookes.

HRDC. 2005. Building on Our Competencies: Canadian Results of the International Literacy and Sills Survey. Ottawa: Statistics Canada.

Justus, M. 2004. “Immigrants in Canada’s Cities.” Pp. 41-49 in Our Diverse Cities, edited by C. Andrew. Ottawa: Canada.

Kuh, G., N. Vesper, M. Connolly, and R. Pace. 1997. College Student Experiences Questionnaire: Revised Norms for the Third Edition. Bloomington, IN: Center for Postseconedary Research and Planning, Indiana University.

Laguian, E., A. Laguian, and T. McGee, eds. 1997. The Silent Debate: Asian Immigration and Racism in Canada. Vancouver: University of British Columbia.

Laureau, A. and E. Weininger. 2003. “Cultural Capital in Educational Research: A Critical Assessment.” Theory and Society 32:567-606.

McMorris, R. and R. Ambrosino. 1973. “Self-Report Predictors: A Reminder.” Journal of Educational Measurement 10:13-7.

Morrison, K. and I. Lui. 2000. “Ideology, Linguistic Capital and the Medium of Instruction in Hong Kong.” Journal of Multilingual and Multicultural Development 21:471-86.

Neuman, S. 2003. Stepping Up: A Framework for Academic Planning at the University of Toronto: 2004-2010. Toronto: University of Toronto.

Ornstein, M. 2006. Ethno-Racial Groups in Toronto 1971-2001: A Demographic Profile. Toronto: Institute for Social Research, York University.

Parcel, T. and E. Menaghan. 1994. Parents’ Jobs and Children’s Lives. New York: Aldine de Gruyter.

Pascarella, E. and P. Terenzini. 1991. How College Affects Students. San Francisco, CA: Jossey-Bass.

Pascarella, E. and P. Terenzini. 2005. How College Affects Students, vol. 2. San Francisco, CA: Jossey Bass.

Pike, G. 1994. “The Relationship between Self-Report and Objective Measures of Student Achievement.” Paper presented at the annual meeting of the Association for the Study of Higher Education, Tuscon, AZ.

Pike, G. 1995a. “Limitations of Using Students’ Self-Reports of Academic Development as Proxies for Traditional Achievement Measures.” Paper presented at the annual meeting of the Association for Institutional Research, Boston, MA.

Pike, G. 1995b. “The Relationship between Self-Reports of College Experiences and Achievement Test Scores.” Research in Higher Education 36:1-21.

Pohlmann, J. and D. Beggs. 1974. “A Study of the Validity of Self- Reported Measures of Academic Growth.” Journal of Educational Measurement 2:115-19.

Portes, A. and R. Schauffler. 1994. “Language and the Second Generations: Bilingualism Yesterday and Today.” The International Migration Review 28:640-61.

Radwanski, G. 1987. Ontario Study of the Relevance of Education and the Issue of Dropouts. Toronto: Ontario Ministry of Education. Robbins, D. 2005. “The Origins, Early Development and Status of Bourdieu’s Concept of ‘Cultural Capital.’” The British Journal of Sociology 56:13-30.

Robson, K. 2003. “Teenage Time Use as Investment in Cultural Capital.” Working Papers of the Institute of Social and Economic Research, University of Essex, Colchester, UK.

Roessingh, H., P. Ehlers, D. Watt, and A. Nettel-Aguirre. 2004. “ESL Students in Tertiary Education Programs: An Exploratory Study of the University of Calgary.” TESOL Quarterly (under review).

Roessingh, H. and P. Kover. 2003. “ESL Learners’ Acquisition of Cognitive Academic Language Proficiency: What Can We Learn from Achievement Measures?” TESL Canada 21:1-21.

Roessingh, H. and D. Watt. 2001. “The Dynamics of ESL Dropout: Plus Ca Change . . . Canadian.” Modern Language Review 58:203-22.

Samuel, J. and K. Bassavarajappa. 2006. “The Visible Minority Population in Canada: A Review of Numbers, Growth, and Labour Force Issues.” Canadian Studies in Population 33: 241-69.

Sullivan, A. 2001. “Cultural Capital and Educational Attainment.” Sociology 35:893-912.

J. PAUL GRAYSON York University

J. Paul Grayson, School of Social Science, Atkinson Faculty, York University, 3700 Keele St, Toronto, Canada M3J 1P3. E-mail: grayson@yorku.ca.

Copyright Canadian Sociology and Anthropology Association, c/o Concordia University Department of Sociology and Anthropology May 2008

(c) 2008 Canadian Review of Sociology and Anthropology, The. Provided by ProQuest LLC. All rights Reserved.