med fmly

Social Psychology of Education 2: 177–197, 1998. © 1998 Kluwer Academic Publishers. Printed in the Netherlands.

177

Family Background, Social and Academic Capital, and Adolescents’ Aspirations: A Mediational Analysis KEVIN MARJORIBANKS University of Adelaide

Abstract. In this study a mediational model was constructed to examine relationships between refined measures of childhood and adolescent social capital and the aspirations of adolescents with varying academic capital and from differing social status and ethnic backgrounds. Longitudinal data were collected from 460 16-year-old Australians and their parents from Anglo-Australian, Greek, and Southern Italian groups. Using partial least squares path modeling the findings suggest: (a) family background, childhood social and academic capital, and adolescents’ social capital combine to have medium to large associations with adolescents’ aspirations, and that the associations are larger for educational aspirations than for occupational aspirations and stronger for males than for females; (b) the mediational model is more successful in explaining family background differences in educational aspirations than variations in occupational aspirations; and (c) after taking into account social and academic capital, Greek males continue to have higher occupational aspirations than do Anglo-Australian and Southern Italian males, and sons of fathers with high social status continue to have higher occupational aspirations than do other male adolescents. The investigation indicates that within encompassing family backgrounds, differences in educational outcomes should be examined in relation to children’s and parents’ perceptions of social and cultural capital and to variations in children’s academic capital.

Adolescents’ idealistic and realistic aspirations have been shown to mediate substantially relationships among family background, individuals’ attributes, and young adults’ social-status attainment (e.g., Carr, Wright & Brody, 1996; Saha, 1994). In social-attainment theory, for example, the intervening process that is hypothesized to account for differences in educational attainment is the development of idealistic value orientations in the form of educational aspirations. On the other hand, in resource-constraint theory, the significant variable mediating relationships between family background and status attainment is the formation of realistic aspirations (Morgan, 1996). Typically, when investigations have examined associations between environments, individuals’ attributes, and adolescents’ aspirations, they have concluded that “socioeconomic background and ability affect aspirations for schooling and careers by way of their realization in school performance and in social support from significant others” (Hauser & Anderson, 1991, p. 246). Many of these investigations have been limited, however, by the use of restricted measures to assess

178

KEVIN MARJORIBANKS

the support provided by significant others. Rumberger (1995) observed “there is considerably less research that has attempted to identify the underlying processes through which family background influences [eventual] success” (p. 587). As a result, our understanding of family background differences in adolescents’ aspirations remains incomplete (Hanson, 1994). The purpose of the present study was to build upon previous research and examine relationships between refined measures of family and school settings and the aspirations of adolescents with varying academic capital and from differing social status and ethnic backgrounds. The conceptual framework for the study was constructed from a bioecological model of development proposed by Bronfenbrenner and Ceci (1994) and from Bourdieu’s (1984, 1988) theory of the educational trajectory of individuals. Bronfenbrenner and Ceci suggested that adolescents’ outcomes are related to proximal processes, which are the enduring forms of interaction that occur in immediate settings such as families and schools and to distal contexts in which the immediate settings are embedded. They claimed the: ... form, power, content, and direction of the proximal processes affecting development vary systematically as a joint function of the characteristics of the developing person, of the environment- both immediate and more remotein which the processes are taking place, and the nature of the developmental outcome under consideration. (p. 572) Furthermore, it is proposed that the affects of the proximal processes on outcomes are more powerful than the influences of distal social contexts, with proximal processes “reducing, or buffering against, environmental differences in developmental outcome.” (p. 574) Bourdieu suggested that to understand the relationships between cultural and social reproduction it is necessary to examine how the “habitus” is produced. The “habitus” is that system of dispositions, such as aspirations, which mediates relations between structures and practice. It is proposed further that the production of the “habitus” is associated with two effects that may either reinforce or offset each other. First, there is an inculcation effect exerted directly by families and schools or by the individuals’ initial social backgrounds. Second, there is a specific effect related to individuals’ predispositions such as the accumulation of academic capital. Bourdieu observed, however, that while individuals are subject to the forces that structure their social space, they may resist “the forces of the field with their specific inertia, that is, their properties, which may exist in embodied form, as dispositions, or in objectified form, in goods, qualifications, etc.” (1984, p. 110). A combination of these theoretical orientations suggests that the influence of initial family background on adolescents’ aspirations is mediated by more immediate family and school settings and by individuals’ academic capital. In the present study, I constructed a mediational model to examine relationships among family background, immediate settings, children’s academic capital, and adolescents’ aspirations.

ENVIRONMENTS AND ASPIRATIONS

179

Environment-Academic Capital Mediational Model FAMILY BACKGROUND

In the model, family background was defined by ethnicity and social status. Entwisle and Astone (1994) proposed that “Ethnic and class differences offer multiple opportunities for comparative research that can enrich our understanding of the key specific issues in human development” (p. 1521). Also, Wentzel (1994) concluded that “parenting practices can vary dramatically as a function of a family’s ethnic and cultural background. Future studies are needed to explore these minority group findings” (p. 286). Family social status was classified separately by fathers’ and mothers’ social position, as studies have revealed that the status of husbands and wives might have differential associations with educational outcomes (e.g., Davies, 1995; Tsai, Gates, & Chiu, 1994). FAMILY BACKGROUND AND SOCIAL CAPITAL

The initial family background is the distal environment in which more immediate settings are embedded and in which individuals’ academic capital is accumulated. Coleman (1988, 1993) indicated the possible nature of the relationship between distal contexts and immediate settings and how that relationship might influence children’s educational development. He proposed that family background is analytically separable into components such as human and social capital. The opportunity of creating a supportive family learning setting is related to parents’ human capital which can be measured approximately by indicators of family background. In contrast, family social capital is generated from the strength of relationships between parents and children. It is the amount and quality of academic interaction between parents and children that provide children with access to parents’ human capital. Coleman (1988) claimed “if the human capital possessed by parents is not complemented by social capital embodied in family relations, it is irrelevant to the child’s educational growth that the parent has a great deal, or small amount, of human capital” ( p. 110). CHILDHOOD SOCIAL CAPITAL

In the present study, a typology devised by Merton (1968, 1976) to investigate associations between social structure and individual behavior was adopted to examine children’s family social capital. One factor in the typology consists of “culturally defined goals, purposes and interests, held out as legitimate objectives for all or for diversely located members of the society ... the prevailing goals comprise a frame of aspirational reference. They are the things ‘worth striving for’ ” (1968, p. 133). The other typological dimension “defines, regulates and controls the acceptable modes of reaching out for these goals” (1968, p. 133).

180

KEVIN MARJORIBANKS

Such a typology suggests that childhood social capital might be defined by parents’ aspirations for their children and by the processes they adopt to realize their aspirations. Similarly, Darling and Steinberg (1993) proposed that, in analyses of family learning environments, three aspects of parenting should be identified. These are parents’ aspirations for their children; parenting style, which provides the emotional context in which parent-child interactions occur; and parents’ involvement in their children’s learning. Parents’ aspirations, style, and involvement were included in the mediational model as measures of childhood social capital. CHILDREN ’ S ACADEMIC CAPITAL

Intellectual ability and academic achievement were included in the model as indicators of the accumulation of academic capital. Teachman (1996) observed that “Intellectual ability is by far the most important predictor of grades and appears to be the funnel through which family background influences academic success” (p. 36). Also, previous investigations have shown children’s cognitive performance to be an important predictor of adolescents’ aspirations (e.g., Hossler & Stage, 1992; Singh & Hernandez-Gantes, 1996). ADOLESCENTS ’ SOCIAL CAPITAL

In the mediational model, more immediate settings were assessed by adolescents’ perceptions of their academic interactions with parents and teachers. Bronfenbrenner and Ceci (1994) indicated the importance of investigating how various settings, within distal social contexts, might be related to individuals’ outcomes. Perceptions of those settings were examined as Wentzel (1994) concluded that adolescents’ outcomes “may be more highly related to their own perceptions of parenting than to what parents think they are doing in the home” (p. 264). Also, Plomin (1995) observed that: ... extant environmental measures are much more passive than active despite the shift from passive to active models of environmental influence. Progress in this field depends on developing measures of the environment that reflect children’s active role in constructing and reconstructing experience. (p. 62)

ADOLESCENTS ’ ASPIRATIONS

Aspirations were chosen for analysis, as they have been shown to mediate substantially the relations between family background and measures of adolescents’ eventual educational and occupational attainments (Saha, 1994). To summarize, the mediational model that I investigated is presented in Figure 1. Variables are organized into blocks which are ordered according to their assumed causal priorities. Relationships among the blocks of predictor variables

ENVIRONMENTS AND ASPIRATIONS

181

Figure 1. Mediational model for relationships between family background, childhood social and academic capital, adolescents’ social capital, and adolescents’ aspirations.

and adolescents’ aspirations were examined using latent variable path modeling techniques. Those variables that are highlighted in the figure represent latent constructs. The manifest indicator variables that were measured in the study and which were considered to form the underlying latent constructs are also shown in the figure. Relationships in the mediational model were examined separately for females and males, as studies have revealed gender-related differences in the formation of aspirations (e.g., Broaded, 1997; Hossler & Stage, 1992; Marini, Fan, Finley, & Beutel, 1996). For the analysis, I hypothesized that (a) adolescents’ social capital would be more strongly related to their aspirations than would family background and measures of childhood social and academic capital, (b) childhood social and academic capital and adolescents’ social capital would mediate relationships between family background and adolescents’ aspirations, and (c) there would be gender-related differences in the nature of the relationships among family background, social and academic capital, and adolescents’ aspirations. Method Sample and Procedure Data were collected from 460 families as part of a longitudinal study. The families lived in Adelaide, the capital city of South Australia, and were selected to reflect the social status distribution of the city. In the first survey, each family had an 11-yearold child, and parents were interviewed in their homes to assess family social status and family social capital in relation to the 11 year olds. Interviews were conducted

182

KEVIN MARJORIBANKS

by experienced social survey interviewers who attended briefing sessions before and during the data collection. The children’s ability and academic achievement were tested by me, in normal classroom settings, to ensure that all students were able to understand the test instructions and to establish, as far as possible, uniform test taking conditions. In a follow-up investigation, five years later, the 16 year olds provided information about their perceptions of their academic interactions with parents and teachers and indicated their educational and occupational aspirations (see Marjoribanks, 1994). The sample included 270 Anglo-Australian, 100 Greek, and 90 Southern Italian families. Greek and Southern Italian families make up two of the largest immigrant groups in Australia (Clifton, 1994). Families were classified as Anglo-Australian if both parents were born in Australia or England and English was the only language spoken in the home. In the Greek and Southern Italian families, both parents had migrated to Australia before or just after the birth of their 11-year-old child, and Greek or Italian was the language typically spoken in their respective homes. Measures FAMILY SOCIAL STATUS

Fathers and mothers reported their levels of educational attainment and occupations during the interviews. Educational levels were assessed on 7-point scales (1 = less than elementary school completed; 7 = postgraduate university education), while parents’ occupational attainments were coded by the Australian National University 3 Scale which is a socioeconomic status index that ranks occupations in the Australian context (Jones, 1989). In the female sample, the correlations between educational attainment and occupations for fathers and mothers were 0.56 and 0.39, respectively. The corresponding correlations in the male sample were 0.54 and 0.40. CHILDHOOD SOCIAL CAPITAL

In the initial survey, a semi-structured interview schedule was developed to measure family social capital. From parents’ responses, factor scales were constructed that assessed parents’ aspirations, parenting style, and parents’ involvement in their 11-year-olds’ learning. Parents’ Aspirations A factor scale of five items, with an alpha reliability estimate of 0.78, assessed the parents’ idealistic and realistic aspirations for their children. The educational aspiration items were “How much education would you really like your child to receive, if at all possible?” and “How much education do you really expect your child to receive?” (1 = leave school as soon as possible; 6 = postgraduate university

ENVIRONMENTS AND ASPIRATIONS

183

education). Parents were also asked “How long have you had these ideas about the amount of education you expect your child to receive?” (1 = just this year; 5 = since the child was born), to gain a sense of the accumulated strength of the aspirations. Similar questions were asked about occupational aspirations: “What kind of job would you really like your child to have when he/she grows up, if at all possible?” and “What kind of job do you really expect your child to have?” These items were coded using the same scale that was adopted to assess parents’ occupations (Jones, 1989). Parents were also asked “How long have you had these ideas about the kind of job you would like your child to have?” The responses to this latter item, however, did not load strongly on the parents’ aspirations factor scale. Parents’ Individualistic-Collectivistic Orientations One of the most significant socialization styles that differentiates between families relates to differences in parents’ individualistic-collectivistic orientations (Kim, Triandis, Kaagitcibasi, Choi, & Yoon, 1994). Individualistic parents tend to socialize their children for self-reliance and independence whereas collectivistic parents orientate children for dependence. Parents’ individualistic-collectivistic orientations, in this study, were assessed by two factor scales. The first scale included seven items that required parents to indicate the age at which they would allow their children to undertake certain activities. Statements were of the form “to sleep at a friend’s place overnight,” “to go on an overnight trip organized by the school,” and “to go to the movies alone.” Response options were combined into 5-point scales (1 = ages 15, 16; 5 = ages 7, 8). High scores reflected a socialization orientation in families in which parents encouraged self-reliance. The scale was labeled as parents’ press for self-reliance and had an alpha reliability estimate of 0.82. In the second scale, parents reacted to five statements such as “Even when a boy/girl gets married his/her main loyalty still belongs to his/her family,” and “When the time comes for a daughter/son to take a job, she/he should try and stay near her/his parents, even if it means giving up good opportunities.” The items were selected from a schedule developed by Strodtbeck (1961), who examined the value orientations of individuals from differing ethnic groups. A 5-point response scale, ranging from 1 (strongly agree) to 5 (strongly disagree) was used to score the items. High scores indicated that parents were sympathetic to their children’s independence from the family, while low scores suggested that parents encouraged dependent behavior by their children. The scale, which had an alpha reliability of 0.84, was labeled parents’ press for independence. Correlations between the measures of these manifest variables of press for selfreliance and independence were 0.53 for females and 0.54 for males. Parents’ Involvement In the analysis, two scales were generated to measure parents’ involvement in their child’s education. The first scale, which was defined as parental teaching, included

184

KEVIN MARJORIBANKS

10 items such as “When your child was small, before starting school, how often did you read to her/him?” (1 = no reading to the child; 6 = just about every day) and “How often would you listen to your child read to you now?” (1 = never listens, 6 = just about every day). In the second scale, which was labeled as parental activeness, there were six items of the form, “How often do you praise your child for work done at school?” (1 = no praise given; 5 = everyday or nearly everyday) and “How much time do you expect your child to devote to homework each school day?” (1 = no time; 5 = more than an hour most days). These two scales had alpha reliability estimates of 0.84 and 0.82, respectively. The correlations between the manifest variables of parental teaching and parental activeness were 0.43 for females and 0.41 for males. ACADEMIC CAPITAL

The 11-year-olds’ academic capital was assessed using measures of intellectual ability and academic achievement. Ability was measured by the Raven Progressive Matrices, normed to the Australian context (Raven, 1989). Academic achievement was defined by children’s mathematics and word performance, which were assessed using standardized tests developed by the Australian Council for Educational Research (ACER). These tests, the Class Achievement Test in Mathematics (Australian Council for Educational Research, 1976) and the Primary Reading Survey Tests (Australian Council for Educational Research, 1976) have reliabilities greater than 0.92. The correlations between mathematics and word performance, for females and males, were 0.50 and 0.48. ADOLESCENTS ’ SOCIAL CAPITAL

In the follow-up study, structured questionnaires were constructed to measure the 16-year-olds’ perceptions of their academic interactions with parents and teachers. Academic Interactions with Parents A family schedule assessed adolescents’ perceptions of their parents’ educational and occupational aspirations for them, the support they had received from their parents in relation to schooling, and their parents’ involvement in their education. Items were of the form “How much education does your father/mother want you to achieve?” (1 = leave school as soon as possible; 7 = graduate from university after postgraduate education), “My father/mother has given me great support while I have been in high school” (1 = strongly disagree; 5 = strongly agree), and “My father/mother often helps me with my homework” (1 = strongly disagree; 5 = strongly agree). From the responses, two 10-item scales were formed that had alpha reliability estimates of 0.84. These scales were designated as adolescents’ perceptions of academic interactions with fathers and mothers. The correlations between these manifest variables were 0.70 and 0.68 for females and males.

ENVIRONMENTS AND ASPIRATIONS

185

Although there was some overlap in the measures of childhood and adolescent family social capital, it is a restriction of the study that parallel measures of the different manifest indicators were not used. The schedules that were constructed for the two surveys do provide, however, refined assessments of a wide range of indicators of the amount and quality of academic interaction between parents and children. Academic Interactions with Teachers Bernstein (1977) proposed that student-teacher interactions might be considered to occur in four interrelated settings which were identified as (a) the regulative setting, which is defined by the nature of authority relationships between teachers and students; (b) the instructional setting, “where the child learns about the objective nature of objects and persons and acquires skills of certain kinds” (p. 481); (c) the imaginative or innovative setting, where students are encouraged to experiment and recreate their world on their terms; and (d) the interpersonal setting, where students are made aware of affective states-their own and those of other students. In the present study, each of the settings was assessed by 5-point Likert-type items such as “Our teachers often discuss with us why the school has certain rules and why they are important,” “Most of my teachers know their subject matter very well and are able to present it in an interesting manner,” “This school is full of teachers with very imaginative and different ways of thinking about things–it is a very exciting place to be,” and “This is a very caring school in which teachers and students care greatly about each other.” A general factor, with an alpha reliability estimate of 0.82, was generated from the responses. The factor, which was labeled as adolescents’ perceptions of academic interactions with teachers, was defined positively by supportive regulative relationships, strong academic teacher orientations, imaginative-academic teaching practices, and caring student-teacher interpersonal relationships. ADOLESCENTS ’ ASPIRATIONS

The adolescents’ idealistic educational and occupational aspirations were assessed by asking them to indicate what educational level and occupation they would really like to achieve, if at all possible, when they were about 25 years old. Realistic aspirations were gauged by asking what educational level they really expected to attain and what job they really expected to have when they were about 25 years old. A 7-point scale (1 = leave school this year; 7 = university degree including postgraduate education) was used to measure the adolescents’ educational aspirations. Occupational aspirations were coded using the same rating scale that was adopted to assess parents’ occupations (Jones, 1989). The correlations between idealistic and realistic educational aspirations were 0.73 and 0.71 for females and males, while the corresponding correlations for occupational aspirations were 0.74 and 0.68.

186

KEVIN MARJORIBANKS

Analysis The relationships among family background, social and academic capital, and aspirations were examined using partial least squares path modeling (Sellin & Keeves, 1997). In this statistical approach, paths are estimated among latent constructs that may be measured by multiple indicators. The limitations of the approach are that it cannot model measurement error or model unmeasured correlations between the manifest variables. However, the advantage of partial least squares modeling for the present analysis is that it does not require stringent distributional assumptions such as observational independence and normality of residuals. Assumptions need not be made about the shape and nature of the underlying distributions of the observed and latent variables. As a result, the approach permits the inclusion of dichotomous variables that are not associated with an underlying continuous distribution such as a set of dummy variables to define ethnic group membership. In this path approach, latent constructs are estimated as linear composites of their manifest indicators, and all information between observed measures is assumed to be conveyed by latent variables. Therefore, in the present mediational path analysis, ethnicity was represented by a set of dummy variables. Each of the other latent constructs in Figure 1 was generated either from a linear composite of their related manifest measures or, in the case of parents’ aspirations, intellectual ability and adolescents’ academic interactions with teachers, reflected in the scores from one manifest indicator. Results A feature of partial least squares path modeling is the explicit estimation of latent variable scores. In Table 1, mean latent scores are presented for female and male adolescents from each ethnic group to provide an initial indication of the nature of group differences in the latent construct measures. The scores have been standardized with means of 50 and standard deviations of 10 to make it simpler to detect possible group patterns in the scores. When the significance of differences in the mean scores were examined using ttests, the findings indicated that Anglo-Australian adolescents were typically from families of higher social status with individualistic-oriented parents who were actively involved in their children’s learning and who expressed moderate aspirations. Greek and Southern Italian adolescents were generally from lower social status families with collectivistic-oriented parents who were less actively involved in their children’s education and who expressed relatively high aspirations. In relation to academic capital, Anglo-Australian females had significantly higher ability scores than did the Southern Italian females while Anglo-Australian males had significantly higher ability scores than did Southern Italian and Greek males. Also, the academic achievement scores of Anglo-Australian children were significantly higher than those of Greek and Southern Italian children. Greek adolescents perceived that their academic interactions with parents were significantly more supportive than were the perceived interactions of Anglo-Australian adolescents.

Latent constructs

Greek C

Southern Italian A–B

Fathers’ status Mothers’ status Parents’ aspirations Parents’ individualistic-collectivistic orientations Parents’ involvement Intellectual ability Academic achievement Adolescents’ academic interactions: with parents with teachers Educational aspirations Occupational aspirations

Females 54.35 54.50 47.73 55.52 54.97 51.48 52.76 49.16 49.81 49.30 49.71

43.06 43.06 56.16 39.09 41.10 49.80 44.27 53.08 49.31 53.46 51.77

Fathers’ status Mothers’ status Parents’ aspirations Parents’ individualistic-collectivistic orientations Parents’ involvement Intellectual ability Academic achievement Adolescents’ academic interactions: with parents with teachers Educational aspirations Occupational aspirations

Males 54.54 55.37 47.54 55.70 55.74 51.24 52.85 49.01 50.72 47.81 49.01

43.32 42.35 53.85 39.41 40.65 47.63 45.34 52.83 49.42 53.80 54.03

∗ p < .05; ∗∗ p < .01; ∗∗∗ p < .001

A

A–C

t-values B–C

42.21 41.28 52.96 41.98 40.71 46.68 43.90 51.37 50.91 51.03 48.87

7.24∗∗∗ 7.70∗∗∗ –4.95∗∗∗ 14.67∗∗∗ 10.94∗∗∗ 0.95 5.02∗∗∗ –2.21∗ 0.27 –2.25∗ –1.16

8.71∗∗∗ 9.73∗∗∗ –3.35∗∗∗ 12.19∗∗∗ 11.44∗∗∗ 2.96∗∗ 5.76∗∗∗ –1.41 –0.65 –1.07 0.47

0.69 1.20 1.53 –1.48 0.25 1.44 0.21 0.71 –0.74 1.17 1.32

40.88 39.39 54.09 46.55 40.81 45.70 45.54 51.22 47.20 48.81 47.04

8.66∗∗∗ 11.50∗∗∗ –4.29∗∗∗ 15.87∗∗∗ 14.28∗∗∗ 2.48∗ 5.21∗∗∗ –2.56∗ 0.85 –4.04∗∗∗ –3.35∗∗∗

6.78∗∗∗ 9.76∗∗∗ –2.96∗∗ 6.11∗∗∗ 9.61∗∗∗ 2.47∗ 3.18∗∗ –1.02 1.61 –0.47 –0.87

1.56 1.74 –0.11 –3.63∗∗∗ 0.09 ‘0.75 –0.10 0.66 0.81 2.10∗ 3.02∗∗

187

Anglo-Australian B

ENVIRONMENTS AND ASPIRATIONS

Table I. Standardized Latent Construct Mean Scores and Significance of Ethnic Group Differences in Mean Scores.

188

KEVIN MARJORIBANKS

The results also indicated that Greek females had higher educational aspirations than did females from Anglo-Australian families while Greek males expressed significantly higher educational aspirations than did Anglo-Australian and Southern Italian males. Greek males had significantly higher occupational aspirations than did males from the other groups. Within ethnic groups, Southern Italian parents expressed more collectivistic orientations for daughters than for sons, and Southern Italian females perceived stronger academic relationships with teachers than did Southern Italian males. While the findings in Table 1 indicated significant ethnic group differences in latent construct mean scores, the results also suggested that these relationships were likely to be confounded by the large ethnic group variations in parents’ social status. That is, these initial results support the proposition that in mediational analyses of family influences on students’ educational outcomes, family background should be assessed by ethnicity and parents’ social status. In Table 2, the findings from the path analysis for educational aspirations are presented. The standardized regression coefficients show the changing effects of the latent construct predictor variables on educational aspirations as the social and academic capital measures were added successively to regression models. In Model 1, the relationships are shown between family background and adolescents’ educational aspirations. Ethnic group membership was represented by a set of dummy variables, and in the regression equations the omitted category was Anglo-Australian. The initial regression coefficients indicated that females from families with high social status fathers and, in particular, if they were from Greek or Southern Italian families had higher educational aspirations than did other female adolescents. When childhood social and academic capital were added to the regression analyses in Model 2, the relationships between ethnicity, fathers’ social status, and the females’ educational aspirations became attenuated and were no longer significant. That is, family background differences in females’ educational aspirations could be attributed to differences in parents’ aspirations, children’s intellectual ability, and academic achievement. The third regression model revealed that the relationships between the measures of academic capital and females’ educational aspirations were no longer significant after taking into account the female adolescents’ perceptions of their academic interactions with parents and teachers. Parents’ earlier aspirations continued to have a significant association with their daughters’ educational aspirations, even after taking into account adolescents’ social capital. For males, the results in Model 1 indicated that fathers’ social status and Greek family membership contributed significantly to higher educational aspiration scores. When childhood social and academic capital were added to the regression analyses in Model 2, the relationships between Greek group membership, fathers’ social status, and educational aspirations were attenuated but remained significant. In Model 3, male adolescents’ perceptions of academic interactions with parents and teachers mediated the relationships between the family background measures, children’s

Predictor variables Family background Greek Southern Italian Fathers’ status Mothers’ status Childhood social and academic capital Parents’ aspirations Parents’ individualisticcollectivistic orientations Parents’ involvement Intellectual ability Academic achievement Adolescents’ social capital Interactions with parents Interactions with teachers Multiple R

Model 1

Females Model 2

Model 3

.27∗∗∗ .21∗∗ .18∗ .09

.15 .15 .10 .02

.09 .04 .04 .01

.32∗∗∗

.24∗∗

Model 1

Males Model 2

Model 3

.39∗∗∗ .13 .27∗∗ .05

.20∗ .02 .16∗ .02

.14 .01 .10 .01

.20∗∗∗

.38∗∗∗

.16∗∗

.05 .03 .16∗ .19∗∗

.04 .01 .07 .11

.01 .20∗ .18∗∗ .22∗∗

.01 .12 .05 .16∗∗

.52∗∗∗

.43∗∗∗ .27∗∗∗ .72∗∗∗

.60∗∗∗

.51∗∗∗ .22∗∗∗ .79∗∗∗

.34∗∗∗

ENVIRONMENTS AND ASPIRATIONS

Table II. Standardized Regression Coefficients for Relationships Between Family Background, Social and Academic Capital, and Adolescents’ Educational Aspirations.

∗ p < .05; ∗∗ p < .01; ∗∗∗ p < .001

189

190

KEVIN MARJORIBANKS

academic capital, and educational aspirations. As for the females in the study, parents’ earlier aspirations for their sons continued to have significant associations with educational aspirations. That is, the findings provided support for the first hypothesis that adolescents’ social capital would be more strongly related to their aspirations than would family background and childhood social and academic capital. Also, the regression models indicated that the initial relationships between family background and adolescents’ educational aspirations were mediated by childhood social and academic capital and adolescents’ social capital. That is, the findings supported the second hypothesis that the relationship between family background and aspirations would be mediated by the intervening measures. Also, the differing patterns of relations in the models in Table 2, between females and males, provided support for the third hypothesis that there would be gender-related differences in the nature of the relationships among the measures in the mediational model. In Table 3, the findings from the path analysis for occupational aspirations are presented. The results in Model 1 show that family background was not related significantly to females’ occupational aspirations. In Model 2, parents’ involvement in their child’s education and children’s academic achievement had significant associations with the occupational aspiration scores. When females’ perceptions of social capital were added to the regression analyses in Model 3, parents’ involvement and academic achievement continued to have significant relationships with occupational aspirations. In contrast, for males, Greek family membership and fathers’ social status had significant initial associations with their occupational aspirations. The results in Model 2 indicated that parents’ aspirations and the measures of children’s academic capital had significant associations with the males’ occupational aspirations. Although these latter associations attenuated the relationships between family background and occupational aspiration scores, the relationships remained significant. In the full regression model, the addition of male adolescents’ perceptions of their academic interactions with parents and teachers mediated the relations between parents’ aspirations and intellectual ability with occupational aspirations. Greek family membership and fathers’ social status, however, continued to have significant associations with the males’ occupational aspirations. The results in Table 3 also revealed that adolescents’ social capital was not consistently more strongly related to occupational aspirations than were the measures of family background and childhood social capital, which indicated that the first hypothesis was not supported (Lubinski & Humphreys, 1990). Similarly, the second hypothesis was not supported as the impact of family background on males’ occupational aspirations was not mediated by the intervening measures. The findings did indicate, however, that there were gender-related differences in the nature of the relationships in the mediational model for adolescents’ occupational aspirations which provided support for the third hypothesis of the study. When effect sizes were calculated (Cohen, 1992), the results revealed that the predictor variables in the mediational path model combined to have large associ-

Predictor variables Family background Greek Southern Italian Fathers’ status Mothers’ status Childhood social and academic capital Parents’ aspirations Parents’ individualisticcollectivistic orientations Parents’ involvement Intellectual ability Academic achievement Adolescents’ social capital Interactions with parents Interactions with teachers Multiple R

Model 1

Females Model 2

Model 3

.16 .10 .08 .14

.17 .12 .01 .05

.19 .12 .01 .04

.11

.18

Model 1

Males Model 2

Model 3

.40∗∗∗ .08 .36∗∗∗ .02

.23∗ .02 .28∗∗∗ .01

.19∗ .01 .23∗∗ .01

.04

.20∗∗

.05

.09 .30∗∗∗ .05 .23∗∗

.04 .26∗∗ .05 .18∗

.15 .09 .15∗ .23∗∗∗

.14 .04 .06 .18∗∗

.3

.38∗∗∗

.27∗∗∗ .13∗ .47∗∗∗

.35∗∗∗ .19∗∗∗ .65∗∗∗

.38∗∗∗

.54∗∗∗

ENVIRONMENTS AND ASPIRATIONS

Table III. Standardized Regression Coefficients for Relationships Between Family Background, Social and Academic Capital, and Adolescents’ Occupational Aspirations.

∗ p < .05; ∗∗ p < .01; ∗∗∗ p < .001

191

192

KEVIN MARJORIBANKS

ations with educational aspiration scores and medium associations with occupational aspirations. Also, the model was more successful in accounting for differences in males’ aspirations than in female aspiration scores. Discussion The findings of the present study replicate, in part, previous investigations of Australian families. Taft (1976) concluded, for example, from three separate investigations that: The educational aspirations of the students of non-English speaking origin are higher than are those of Australians, especially in the working class... The children of immigrants of English speaking countries have comparatively low expectations even though many of these children have been in Australia for most or all of their life. (p. 306) In a study involving senior high school students, Meade (1983) indicated that parents’ and children’s aspirations in relation to university attendance were greater in Greek families than in Anglo-Australian and Italian families. Saha (1985), in a further investigation of senior high school students, revealed that male students who were classified as European had significantly higher preferred and expected occupations than did Australian and British/Irish students. Clifton, Williams and Clancy (1991) constructed a social-psychological model to examine ethnic group differences in academic attainment. They revealed “Even though the Greek and Italian students have lower academic achievement than have the Australian and English students, their social-psychological support for education is generally higher” (p. 123). Furthermore, they concluded that social-psychological variables “do not mediate or suppress all the effects of ethnicity on the completion of secondary school ... we must conclude that the effects of ethnicity within the academic attainment process are considerably more complicated than was previously assumed” (p. 124). An implication of the earlier investigations and of the present study is that parents and students from certain family backgrounds are likely to become dissatisfied and disenchanted with schools if their high aspirations are not translated into academic success such as university participation. Taft (1976) predicted, however, that the high aspirations of immigrant parents and children will be far from satisfied. He suggested “Even if the economy could provide the necessary number of openings, an important proportion of students will be blocked by the various competitive screenings in the mobility process” (p. 307). Although it has been suggested by Williams, Long, Carpenter, and Hayden (1993) that there is some evidence that young Australians born in Greece and Italy have lower than overall university participation rates, Lamb, Polesel and Teese (1995) observed that national studies are not available to indicate with precision the numbers of students in Australian universities from differing ethnic groups. What is required now are Australian-wide investigations that extend the mediational model to include school

ENVIRONMENTS AND ASPIRATIONS

193

completion and university participation rates. Such studies should examine to what extent students from different family backgrounds are blocked, as Taft predicted, by various competitive screenings in the mobility process, and if such “blocking” exists, whether it affects parents’ and students’ perceptions of the fairness and equity of the educational system. A further implication of the present study relates to the strong associations that were found between adolescents’ perceptions of their academic interactions with parents and teachers and their aspirations. It is generally agreed that students’ school-related outcomes will be enhanced when family and school learning situations are supportive of learning and in harmony with each other. Kellaghan, Sloane, Alvarez, and Bloom (1993) suggested: ... when the home and school have divergent approaches to life and to learning, children are likely to suffer in their school learning. Conversely, when home and school have similar emphases on motivation and learning, children are likely to do well. (p. 145) Similarly, Wang, Haertel, and Walberg (1993) proposed “the home is central to students’ daily experience. Consequently, the home functions as the most salient out-of-school context for student learning, amplifying or diminishing the school’s effect on school learning” (p. 278). In general, the current study indicated that adolescents’ aspirations were enhanced when positive perceptions of their academic interactions with parents and teachers were in harmony. Such a finding suggests that while it is important for teachers to become involved with parents in partnership programs, it is also important to involve children in those activities so that their perceptions of their immediate learning settings can be explored. That is, educational programs designed to enrich school outcomes need to investigate family and school social capital and, at the same time, examine the interpretations and meanings that children and adolescents have of that social capital. Unless parentinvolvement programs adopt strategies that explore students’, as well as parents’ and teachers’ perceptions of social capital, then it is unlikely that the impact on school outcomes of altering immediate settings will be fully realized. The present analysis of Australian adolescents suggests the propositions that (a) family background, childhood social and academic capital, and adolescents’ social capital combine to have medium to large associations with adolescents’ aspirations and that the associations are larger for educational aspirations than for occupational aspirations and stronger for males than for females; (b) the mediational model is more successful in explaining family background differences in educational aspirations than variations in occupational aspirations; and (c) after taking into account social and academic capital, Greek males continue to have higher occupational aspirations than do Anglo-Australian and Southern Italian males, and sons of fathers with high social status continue to have higher occupational aspirations than do other male adolescents. The third proposition suggests a number of possible refinements of the theoretical framework developed for the present study. First, the introduction of new

194

KEVIN MARJORIBANKS

variables, or more refined measures of the current intervening variables, might account for the family background differences in males’ occupational aspirations. In this study, parents’ individualistic-collectivistic orientations were not associated with aspirations. Authoritative parenting style, that is defined by parental warmth, behavior supervision and strictness, and psychological autonomy granting and which has been shown to be related to academic success, might be a more appropriate measure of parents’ orientations (Fletcher, Darling, Steinberg, & Dornbusch, 1995; Smetana, 1995). Also, adolescents’ perceptions of friends’ experiences and behaviors are likely to influence their own aspirations and could profitably be included in the mediational model. Second, the persistence of significant residual effects of ethnicity and fathers’ social status on males’ occupational aspirations may reflect differences in the cultural value placed on education by parents from different social backgrounds. Bourdieu (1984) argued, for example, that families from certain social backgrounds have cultural capital that is elaborated in a “taste” for education and reflected in an emphasis on the intrinsic value of scholastic success (also see Fejgin, 1995). Wells and Serna (1996) indicated “academic qualifications and high-status educational titles are to cultural capital what money and property titles are to economic capital” (p. 97). The transmission of cultural capital in families may not be captured by the measures which are used in the present study to assess family social capital. Instead, such cultural capital might more appropriately be measured by parents’ life styles, their behaviors and conventions, and the establishment of cultural codes which continually communicate the value and importance of education for the well being of individuals and for the betterment of society (Archer & Blau, 1993). That is, in an elaborated mediational model, a measure of cultural capital should be included but which is expanded from its traditional and more limited emphasis on socialization into high status culture (Kalmijn & Kraaykamp, 1996). The present study has indicated that students’ school outcomes need to be considered as being embedded in backgrounds that are defined by the complementarity, contradictions, and tensions of gender, ethnicity and parents’ social status (McCarthy, 1988). Within these encompassing backgrounds, differences in outcomes should be examined in relation to children’s, parents’ and teachers’ perceptions of social and cultural capital and to variations in children’s academic capital. Investigations should now examine the elaborated mediational model for different outcomes, with more refined measures of the intervening variables, and for students from other social backgrounds. Such research will enable us to increase our theoretical understanding of how the human, cultural and social capital of families and schools interact with academic capital to affect students’ aspirations and their eventual educational outcomes.

ENVIRONMENTS AND ASPIRATIONS

195

References Archer, Melanie & Blau, Judith R. (1993). Class formation in the nineteenth century: The case of the middle class. In J. Blake & J. Hagen (Eds.), Annual review of sociology (Vol. 19). Palo Alto, CA: Annual Reviews, pp. 17–41. Australian Council for Educational Research (1976). Classroom achievement test in mathematics. Melbourne: Author. Australian Council for Educational Research (1976). Primary reading survey tests. Melbourne: Author. Bernstein, Basil B. (1977). Class, codes and control. Towards a theory of educational transmissions (Vol. 3). London: Routledge & Kegan Paul. Bourdieu, Pierre (1984). Distinction: A social critique of the judgement of taste. London: Routledge & Kegan Paul. Bourdieu, Pierre (1988). Homo academicus. Cambridge, U.K.: Polity Press. Broaded, C. Montgomery (1997). The limits and possibilities of tracking: Some evidence from Taiwan. Sociology of Education, 70, 36–53. Bronfenbrenner, Urie & Ceci, Stephen J. (1994). Nature-nurture reconceptualization in development perspective: A bioecological model. Psychological Review, 101, 568–586. Carr, Rhoda V., Wright, James D., & Brody, Charles J. (1996). Effects of high school work experience a decade later: Evidence from the National Longitudinal Survey. Sociology of Education, 69, 66–81. Clifton, Rodney A. (1994). Race and ethnicity in education. In T. Husén & T. N. Postlethwaite (Eds.), The international encyclopedia of education. Oxford, U.K.: Pergamon, pp. 4891–4896. Clifton, Rodney A., Williams, Trevor, & Clancy, Jeff (1991). The academic attainmentof ethnic groups in Australia: A social psychological model. Sociology of Education, 64, 111–126. Cohen, Jacob (1992). A power primer. Psychological Bulletin, 112, 155–159. Coleman, James S. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, 95–120. Coleman, James S. (1993). The rational reconstruction of society. American Sociological Review, 58, 1–15. Darling, Nancy & Steinberg, Laurence (1993). Parenting style as context: An integrative model. Psychological Bulletin, 113, 487–496. Davies, Scott (1995). Reproduction and resistance in Canadian high schools: An empirical analysis. British Journal of Sociology, 46, 662–687. Entwisle, Doris R. & Astone, Nan M. (1994). Some practical guidelines for measuring youth’s race/ethnicity and socioeconomic status. Child Development, 65, 1520–1540. Fejgin, Naomi (1995). Factors contributing to the academic excellence of American Jewish and Asian students. Sociology of Education, 68, 18–30. Fletcher, Anne C., Darling, Nancy E., Steinberg, Laurence, & Dornbusch, Sanford M.(1995). The company they keep: Relation of adolescents’ adjustment and behavior to their friends’ perceptions of authoritative parenting in the social network. Developmental Psychology, 31, 300–310. Hanson, Sandra L. (1994). Lost talent: Unrealized educational aspirations and expectations among U.S. youths. Sociology of Education, 67, 159–183. Hauser, Robert M. & Anderson, Douglas K. (1991). Post-high school plans and aspirations of Black and White high school seniors: 1976–86. Sociology of Education, 64, 263–277. Hossler, Don & Stage, Frances K. (1992). Family and high school experience influences on the postsecondary educational plans of ninth-grade students. American Educational Research Journal, 29, 425–451. Jones, F. Lancaster (1989). Occupational prestige in Australia: A new scale. Australian and New Zealand Journal of Sociology, 25, 187–199.

196

KEVIN MARJORIBANKS

Kalmijn, Matthijs & Kraaykamp, Gerbert (1996). Race, cultural capital, and schooling:An analysis of trends in the United States. Sociology of Education, 69, 22–34. Kellaghan, Thomas, Sloane, K., Alvarez, B., & Bloom, Benjamin S. (1993). The home environment and school learning: Promoting parental involvement in the education of children. San Francisco: Jossey-Bass. Kim, U., Triandis, H. C., Kagitcibasi, C., Choi, S-C., & Yoon, G. (1994). Individualism and collectivism: Theory, method and applications. Newbury Park, CA: Sage. Lamb, Stephen, Polesel, John, & Teese, Richard (1995). Where do they go? An evaluation of sources of data used for the monitoring of students’ destinations and other educational outcomes in Australia. Canberra: Australian Government Printing Service. Lubinski, David & Humphreys, Lloyd G. (1990). Assessing spurious “moderator effects”: Illustrated substantively with the hypothesized (“synergistic”) relation between spatial and mathematical ability. Psychological Bulletin, 107, 385–393. Marini, Margaret M., Fan, Pi-Ling Fan, Finley, Erica, & Beutel, Ann M. (1996). Gender and job values. Sociology of Education, 69, 49–65. Marjoribanks, Kevin (1994). Families, schools and children’s learning: A study of children’s learning environments. International Journal of Educational Research, 21, 439–555. McCarthy, Cameron (1988). Rethinking liberal and radical perspectives on racial inequality in schooling: Making the case for nonsynchrony. Harvard Educational Review, 58, 265–279. Meade, Phil (1983). The educational experience of Sydney high school students.Canberra: Australian Government Printing Service. Merton, Robert K. (1968). Social theory and social structure. New York: Free Press. Merton, Robert K. (1976). Sociological ambivalence and other essays. New York:Free Press. Morgan, Stephen L. (1996). Trends in Black-White differences in educational expectations: 1980–92. Sociology of Education, 69, 308–319. Plomin, Robert (1995). Genetics and children’s experiences in the family. Journal of Child Psychology and Psychiatry, 36, 33–68. Raven, J. (1989). The Raven Progressive Matrices: A review of national norming studies and ethnic and socioeconomic variation in the United States. Journal of Educational Measurement, 26, 1– 16. Rumberger, Russell W. (1995). Dropping out of middle school: A multilevel analysis of students and schools. American Educational Research Journal, 32, 583–625. Saha, Lawrence J. (1985). The legitimacy of early school learning: Occupational orientations, vocational training plans, and educational attainment among urban Australian youth. Sociology of Education, 58, 228–240. Saha, Lawrence J. (1994). Aspirations and expectations of students. In T. Husén & T. N. Postlethwaite (Eds.), International encyclopedia of education. Oxford, U.K.: Pergamon, pp. 354–358. Sellin, Norbert & Keeves, John P. (1997). Path analysis with latent variables. In J. P.Keeves (Ed.), Educational research, methodology, and measurement: An international handbook (2nd. ed.). Oxford, U.K.: Pergamon, pp. 633–640. Singh, Kusum & Hernandez-Gantes, Victor M. (1996). The relation of English language proficiency to educational aspirations of Mexican-American eighth graders. Journal of Early Adolescence, 16, 253–273. Smetana, Judith G. (1995). Parenting styles and conceptions of parental authority during adolescence. Child Development, 66, 299–316. Strodtbeck, Fred L. (1961). Family integration, values and achievement. In A. H.Halsey, J. Floud, & C. A. Anderson (Eds.), Education, economy and society. New York: Free Press, pp. 315–347. Taft, Ronald (1976). Secondary school children of immigrant families in Victoria. In R.K. Browne & D. J. Magin (Eds.), Sociology of education. Melbourne: Macmillan, pp. 303–307. Teachman, Jay D. (1996). Intellectual skill and academic performance: Do families bias the relationship? Sociology of Education, 69, 35–48.

ENVIRONMENTS AND ASPIRATIONS

197

Tsai, Shu-Ling, Gates, Hill, & Chiu, Hei-Yuan (1994). Schooling Taiwan’s women:Educational attainment in the mid-20th century. Sociology of Education, 67, 243–263. Wang, Margaret C., Haertel, Geneva D., & Walberg, Herbert J. (1993). Toward a knowledge theory of school learning. Review of Educational Research, 63, 249–294. Wells, Amy S. & Serna, Irene (1996). The politics of culture: Understanding local political resistance to detracking in racially mixed schools. Harvard Educational Review, 66, 93–118. Wentzel, Kathryn R. (1994). Family functioning and academic achievement in middle school: A socio-emotional perspective. Journal of Early Adolescence, 14, 268–291. Williams, Trevor, Long, Michael, Carpenter, Peter, & Hayden, Martin (1993). Entering higher education in the 1980s. Canberra: Australian Government Publishing Service.

Biographical Note Kevin Marjoribanks is professor of education and head of the Graduate School of Education, The University of Adelaide, Australia. From 1987 to 1994 he was Vice-Chancellor/President of The University of Adelaide. His research interests relate to an examination of the associations between family and school learning environments and students’ school-related outcomes.