PARENTAL INVOLVEMENT WITH ADOLESCENTS' EDUCATION: DO DAUGHTERS OR SONS GET MORE HELP?
This research examined whether parents were involved differently with the education of their adolescent daughters and sons. The investigation used data from the National Education Longitudinal Study (NELS), which collected information from approximately 25,000 eighth-grade students. Several types of parental involvement were analyzed for gender differences, including school discussion, parent-school connection, parental expectations, parental attendance at school events, and three measures of parental supervision (checking homework, limiting television watching, and limiting going out with friends). The results showed that, net of students' grades, tests scores, and educational aspirations, parents helped daughters in some ways and sons in other ways. Generally, daughters experienced more parental involvement with their education than did sons. The findings are discussed in terms of parents' traditional socialization practices versus a shift in parental treatment in response to social trends.
Parental involvement with children's education has been the subject of research for several decades, and this topic continues to be of interest (recent studies include Bogenschneider, 1997; Eccles, Jacobs, & Harold, 1990; Epstein, 1991; Muller, 1998; Schneider & Coleman, 1993; Smith, 1992; Snow, Barnes, Chandler, Goodman, & Hemptill, 1991; Teachman, Paasch, & Carver, 1996; Useem, 1992). Generally, studies have found a positive relationship between the overall level of involvement by parents and the academic performance of students. While much of the research has focused on the early years of schooling, recent studies indicate that parental involvement is also important for older students. For example, researchers reported that parental involvement positively affected the grades and mathematics test scores of adolescent students (Muller, 1993, 1998), decreased the odds of a student dropping out of high school (Teachman et al., 1996), had positive effects on the grades of high school seniors and the amount of time they devoted to homework (Fehrmann, Keith, & Reimers, 1987), and contributed to successful placement of students in higher ability mathematics groups (Useem, 1992).
In short, research shows that parental involvement with children's education is important for positive academic experiences and successful outcomes. Yet, there is little information about whether parents' involvement differs for daughters and sons. Some of the literature suggests that traditional socialization practices result in parents shortchanging daughters (Smith, 1992; Wellesley College Center for Research on Women, 1992; Wigfield & Eccles, 1994). Thus, investigating differences in parental involvement is important because it could help explain gender differences in the educational experiences of adolescents, and contribute to knowledge about conditions that foster gender stratification in work outcomes.
The present research examined whether parents were involved differently with the education of their adolescent daughters and sons. The study used data from the National Education Longitudinal Study (NELS), which collected information from approximately 25,000 eighth-grade students. Several types of parental involvement were analyzed for gender differences, including school discussion, parent-school connection, parental expectations, parental supervision, and parental attendance at school events.
Gender equality has achieved greater acceptance in society, and one would expect parents to be more likely to treat their female and male children equally. Nevertheless, the results of recent studies suggest that parents favor sons over daughters in various ways. For example, researchers have reported that fathers who have sons are more involved with their children (Harris & Morgan, 1991), mothers of sons are more concerned about child obedience and the possible negative effects of their own employment (Downey, Jackson, & Powell, 1994), and parents of sons are less likely to divorce (Morgan, Lye, & Condran, 1988).
The literature on gender role socialization maintains that parents treat daughters and sons differently due to the patriarchal values of society, which elevate males over females in the social order (Lorber, 1994; Wellesley College Center for Research on Women, 1992). Research has found that traditional socialization practices have a male bias; sons, as compared with daughters, are provided with greater opportunities for personal autonomy and achievement (Eccies et al., 1990; Entwisle, Alexander, & Olson, 1994; Saltiel, 1985). Specifically, studies have linked gender differences in outcomes to socialization that traditionally has emphasized personal relationships, dependency, conformity, and submissiveness for females, versus personal achievement, autonomy, and assertiveness for males (see Block, 1983, and Marini & Brinton, 1984, for reviews). Further, females experience a reduction in self-esteem during adolescence, negatively affecting their aspirations and attainments (Smith, 1992; Wigfield & Eccles, 1994 ). This body of research suggests that parents may be involved with daughters and sons in ways that produce important gender differences.
Gender differences have been reported in regard to students' skills and participation in math and science courses (Catsambis, 1994; Entwisle et al., 1994), perceptions of academic ability (Wigfield & Eccles, 1994), educational expectations (Hanson, 1994), and parents' perception of students' math ability (Eccles et al., 1990; Eccles & Jacobs, 1986). Empirical evidence suggests that differential treatment of females and males concerning academics increases in the higher grades (Catsambis, 1994; Entwisle et al., 1994). Furthermore, parental expectations that males will do better in math and science and the belief that these courses are more difficult for females appear to be independent of actual academic behavior (Eccles et al., 1990). The negative consequences of this gender bias have been demonstrated; female students, compared with male students, have been found to have lower selfconcepts concerning their mathematical ability (Wigfield & Eccles, 1994). Catsambis (1994) found that the proportion of male stu dents expressing aspirations in math-related fields was more than double that of female students. Hanson (1994) reported that among high school seniors who showed early signs of talent, females were more likely than their male counterparts to aspire to a college degree, but not to believe that they would attain one.
On the one hand, the evidence regarding adolescent educational aspirations, experiences, and attainments suggests that gender role socialization has detrimental effects for females. On the other hand, recent reports indicate that females earn higher grades, are slightly more likely to graduate from high school, and enter and graduate from college at about the same rate as their male counterparts (Mare, 1995; U.S. Department of Education, 1996). Yet, a major incentive for success in school is later occupational rewards, and women generally have lower status in the labor market than do men (Spain & Bianchi, 1996). Mickelson (1989) suggests that the anomaly of female students doing well academically while facing a future of lower occupational status might be explained by gender role socialization that stresses "good student" behavior on the part of females (e.g., being obedient, pleasing others). Additionally, researchers point out that female students are less likely to enroll in higher level math and science courses, and such educational choices contribute to differences in occupational outcomes (Catsambis, 1994; Entwisle et al., 1994; Lorber, 1994). Thus, most of the evidence suggests that differential gender socialization is detrimental to females; however, its severity is unclear given the recent increase in gender equality in educational attainment.
The literature indicates that parents contribute to the education of their children in various ways. For example, effective parental behaviors include helping children with their homework, encouraging them to study, answering questions, offering guidance on educational decisions, having contact with the school and teachers, and attending school events (Fehrmann et al., 1987; Schneider & Coleman, 1993; Snow et al., 1991; Sui-Chu & Willms, 1996; U.S. Department of Education, 1987). In one of the few studies on parental involvement focusing on gender differences, Bogenschneider (1997), using data from students attending nine schools in California and Wisconsin, found that involvement of fathers did not differ by gender of the child, but mothers were more involved with daughters than with sons. The present study extended prior research by using a nationally representative sample of eighth-grade students to investigate whether several types of parental involvement differ for sons and daughters, such as parent-chi ld discussion, parent-school connectivity, and parental supervision, attendance, and expectations.
Recent studies have shown that parent-child conversations concerning school-related topics contribute to educational success, and daughters talk more with their parents about school matters than do sons (Muller, 1993, 1998; Sui-Chu & Willms, 1996). However, previous analyses did not control for academic factors (e.g., grades, test scores, educational aspirations), and it is possible that parents talk more frequently with good students--regardless of gender--and fes earn higher grades and are generally better students than are males. These academic factors were controlled in the present study. Further, the literature on gender socialization suggests that females are taught to be more dependent and focused on others. They may therefore be more likely to engage in discussions with their parents, regardless of their academic standing. Thus, it was hypothesized that, net of academic factors, daughters would talk more with their parents about educational matters than would sons.
Also considered were the frequency of parents' attendance at school events in which their child participated (parental attendance) and the interaction of parents with the school (parent-school connectivity), such as going to school meetings, having direct contact with the teacher, and visiting the classroom. Both types of parental involvement are important contributors to the educational achievement of students (Muller, 1993; Muller & Kerbow, 1993; Useem, 1992). Studies have found that parents communicate with the school more on behalf of sons than daughters (Sui-Chu & Willms, 1996; Muller, 1998). However, prior analyses did not control for the reason behind the communication: positive (e.g., discussing placement in advanced courses) or negative (e.g., discussing behavioral problems). The literature regarding socialization practices suggests greater involvement on behalf of sons. Thus, it was hypothesized that parents would be more connected with the school and attend more events for sons than for daughters, net of academic factors.
Parents' decisions and guidance on how adolescents spend their free time, and the importance placed on completing homework, indicate the level of family regulation. These factors have been shown to affect educational outcomes (Muller & Kerbow, 1993; Fehrmann et al., 1987). To examine this type of parental involvement, three measures of parental supervision were used: checking homework, limiting television watching, and limiting going out with friends. Research has shown that parents supervise daughters more closely than they do sons (Block, 1983; Muller, 1998). This fits the notion that females are socialized to be dependent and obedient, while males are socialized to be independent and self-willed. In accord with traditional gender socialization, it was hypothesized that, net of academic factors, daughters would be more strictly supervised than would sons.
Lastly, parental involvement can also take the form of expectations for educational achievement. Students' perceptions of parental expectations have important effects on educational outcomes (Muller & Kerbow, 1993). Males have traditionally attained higher educational and occupational status than have females, with studies showing that parents have higher expectations for sons than for daughters (Eccles et al., 1990; Marini & Brinton, 1984). Thus, it was hypothesized that, net of academic factors, parents would have higher educational expectations for sons than for daughters.
In sum, parental involvement is important for educational achievement, and the present research tested for differences in parents' involvement with daughters and sons. Based on the literature, the general hypothesis was that, net of academic factors, parents would be more involved in the education of their adolescent sons. However, school discussion and parental supervision would likely be exceptions, with daughters receiving more of these types of parental involvement.
The data for this study came from the National Education Longitudinal Study (NELS), which was funded by the National Center for Education Statistics (1994). NELS, using a nationally representative sample of students in public and private schools, provides many detailed measures of parents' involvement with the education of their adolescent children. These data were gathered using a two-stage probability sampling design that first randomly selected 1,052 United States schools. Second, approximately 25,000 eighth-grade students from the selected schools were sampled.
Student information from the 1988 wave was used to investigate parental involvement while the adolescent was in the eighth grade. Parental involvement was thus analyzed from the adolescents' perspective, drawing upon the sociological idea that if an individual defines a situation as real, it is real in its consequences (Thomas & Thomas, 1928). In short, the students' accounts were considered important indicators of how they experienced parental involvement with their education.
Parental involvement with adolescents' education (type of involvement and amount of time spent) was represented by the following dependent variables: school discussion, parental expectations, parent-school connection, parental supervision, and parental attendance. Indexes were constructed for school discussion, parent-school connection, and parental expectations. Factor analysis was used for confirmation of the indexes, and reliability was checked by calculating Cronbach's alpha coefficients.
School discussion (alpha = .62). This index is the sum of five questions on the frequency of student discussion with parents regarding: (1) selection of courses or programs at school, (2) school activities or events, (3) things studied in class, (4) planning high school program (talked with mother), and (5) planning high school program (talked with father). Responses were coded 0 = not at all and 1 = once or more.
Parent-school connection (alpha = .49). This index is the sum of three questions that asked students to indicate whether parents: (1) attended a school meeting, (2) phoned or spoke with a teacher or counselor, and (3) visited the classroom. Responses were coded 0 = no and 1 yes.
Parental expectations (alpha = .91). This index is the sum of two questions regarding mother's and father's educational expectations for the student. Responses were coded as follows: 0 = less than high school, 1 = graduate from high school, 2 = some college, 3 = graduate from college, and 4 = attend graduate school.
Parental supervision. Three questions asked students to indicate parents' level of supervision: (1) parents check homework, (2) parents limit television watching, and (3) parents limit going out. Responses were coded 0 = never or rarely and 1 = sometimes or often.
Parental attendance. One question asked whether parents attended a school event in which the student participated. Responses were coded 0 = no and 1 = yes.
Factor analysis indicated a low alpha for the three questions about parental supervision, suggesting that these questions measured distinct aspects of parental involvement and not different dimensions of the same construct. Hence, they were treated as separate dependent variables in the analysis. Table 1 shows the weighted and unweighted means of the dependent variables and the key independent variables used in the analysis.
Research has revealed gender differences in grades and test scores, and these, along with educational aspirations, have been found to be associated with parental involvement (Schneider & Coleman, 1993; Saltiel, 1985; Cohen, 1987). However, the relationship between academic performance and parental involvement is not clear. It is possible that parents are more interactive with daughters and sons who do well in school and have high educational aspirations because these students appreciate and request the extra attention. Alternatively, parents might be more involved with those adolescents who are in trouble academically, or who have low aspirations, in an effort to improve their educational performance. Therefore, the analysis included academic factors (i.e., grades, test scores, and educational aspirations) to control for the effects of students' academic performance and educational orientation on parental involvement.
Grades -- a composite measure of students' self-reports about their grades in English, mathematics, science, and social studies--could range from .5 to 4.0. Test scores is a composite compiled from scores on reading and mathematics achievement tests. Educational aspirations was derived from a question that asked students how far they thought they would get in school, with responses ranging from 1 = less than high school to 4 = postgraduate work. Table 1 shows the mean scores for these independent variables.
The following background variables may be related to parental involvement and were included in the analysis as controls, but the effects were not analyzed: race/ethnicity, family type, nativity, region, urbanicity, school type, number of siblings, and socioeconomic status.
Regression models, with the parental involvement measures as dependent variables, are shown in Tables 2 and 3. Weighted data were used in all of the analyses, and unstandardized coefficients are reported. First, each parental involvement measure was regressed on gender (a dummy variable with female contrasted to male) along with background variables. Next, grades, test scores, and educational aspirations were added to the first regression model to assess the effects of gender on parental involvement, net of academic factors. Ordinary least squares (OLS) regression was used for the three dependent variables that were continuous measures, and logistic regression for the four dependent variables that were dichotomous.
Model 1 (see Table 2) shows that female students engaged more frequently in school discussion with parents than did male students. This finding could be interpreted as supporting the literature on traditional gender socialization, which suggests that daughters are more dependent upon their parents than are sons. Alternatively, it may indicate that parents are very engaged in the school life of their daughters. Thus, with these data, it is not possible to determine the causal direction.
Model 2 controlled for grades, test scores, and educational aspirations. As expected, there was a reduction in the magnitude of the gender effect for school discussion. Regardless, the positive and significant effect for females remained, showing, in support of the hypothesis, that daughters engaged more frequently in discussion of educational matters with their parents than did sons.
Contrary to the hypothesis regarding parental expectations, Model 1 shows that parents had higher expectations for the educational attainment of daughters, and the effect remained in Model 2, which controlled for academic factors. However, the reduction in the gender coefficient from Model 1 to Model 2 was substantial; test scores and educational aspirations largely explained the gender difference in parental expectations. Nevertheless, the effect remained positive and significant, and one possible explanation is that parents had higher educational expectations for daughters due to current conditions in the labor market, in which postsecondary education is necessary for females to get higher paying jobs.
Regarding the effect of gender on parent-school connection, Model 1 supports previous research findings and the hypothesis that parents are more involved with the school on behalf of sons. The effect remained in Model 2, and academic factors did not appear to explain the gender difference. It remains unclear as to the degree to which this parental involvement pertained to nonacademic issues, such as behavioral problems, which perhaps are more common for adolescent males. However, if positive parent-school connection was tapped, as the research on tracking placement suggests (Useem, 1992), then daughters appear to be at a disadvantage.
Table 3 presents the logistic regressions of the parental supervision variables. The coefficient for gender in Model 1 shows that parents were less likely to check the homework of daughters as compared with sons. After controlling for the academic factors, namely grades, test scores, and educational aspirations (Model 2), the effect for gender was slightly lower. Prior research suggests that females are socialized to be good students, and perhaps the present finding indicates that daughters, as compared with sons, more responsibly complete homework, and consequently parents do not feel as great a need to check on them.
There was no gender difference in terms of limiting the amount of time students were allowed to watch television. Model 2 controlled for academic factors, and there was little effect on the coefficient for gender. Previous studies have shown that parental involvement regarding limits on television watching is important for successful educational outcomes (Fehrmann et al., 1987), and it appears here that parents placed limits equally for daughters and sons.
Confirming the hypothesis, females were more likely than males to have their parents limit the amount of time spent socializing with friends (Model 1). This is in line with traditional socialization practices that allow sons more autonomy. Model 2 shows that academic factors explained some of this gender difference, but the effect remained significant and positive. It is possible that this gender difference reflects parents' greater efforts to protect daughters.
Of the three parental supervision variables, placing restrictions on going out was the only one supporting the hypothesis that daughters are more closely supervised than are sons. Daughters were significantly less likely to have parents check their homework, and there was no gender difference in the likelihood of parents limiting students' time watching television. The different effects show why these measures could not be grouped into one index, and indicate that parental supervision is indeed a multifaceted aspect of parental involvement.
Contrary to the hypothesis, the coefficient for gender in Model 1 shows that parental attendance at events in which the student participated was more likely for daughters than for sons. Controlling for academic factors (Model 2) substantially reduced the magnitude of the coefficient for gender. In fact, over half of the gender difference in parental attendance was explained by academic factors, indicating that females being better students contributed to parents attending their school events.
Overall, test scores and educational aspirations were important predictors in most of the models. Interestingly, grades showed significant effects only in the models for school discussion and parental attendance. Clearly, these academic factors explain gender differences in parental involvement, but only in part.
Previous studies have shown that parental involvement is a key factor in the academic success of students. The present research sheds light on important gender differences regarding parental treatment of adolescents. Three of the hypotheses about parents treating daughters and sons differently were supported, which suggests that parents help daughters in some ways and sons in other ways. However, in contrast to the general hypothesis, the results showed that daughters received more attention from their parents than did sons on four of seven measures of involvement, net of academic factors.
These results raise questions about what the gender differences in parental involvement indicate. First, the parental involvement measures did not identify parents' motivation for treating daughters and sons differently. Consequently, there are several ways to interpret the findings. Perhaps parents are more involved with daughters because there is a greater emphasis on educational attainment for females due to current social conditions, such as delayed marriage and more divorce, which require females to be capable of supporting themselves rather than relying on a husband. These types of social changes may have altered traditional socialization practices that favored males. It is also possible that parents currently socialize their daughters in ways that reflect the lessened social stigma surrounding female employment and success. Perhaps with more women in the U.S. labor force than any time previously, the higher educational expectations for daughters are due to the stronger influence of working mothers. Th us, simply stated, high aspirations and achievement are now considered beneficial for females as well as males.
On the other hand, the findings may be interpreted as evidence of the persistence of traditional gender socialization. That is, the greater involvement of parents with their daughters, compared with sons, may show that adolescent females are more dependent upon others than are adolescent males.
Another possibility is that the findings reveal a reciprocal relationship between daughters and their parents. Daughters might report greater parental involvement because they have more positive experiences with parents, whereas adolescents might be less inclined to report parental involvement if the activities were negative, critical, or punitive (Felson & Zielinski, 1989). In addition, if adolescent females are more obedient and cooperative than are adolescent males, that might influence the nature of the parental involvement and affect student responses.
These are plausible interpretations of the findings. However, definitive explanations of the conditions that foster gender stratification are beyond the scope of this study.
Rather than confirming that gender inequality is fostered by parents' encouragement of, and support for, sons' achievement, the results indicate that, net of academic factors, parents are more involved with daughters. This raises questions about the relationship between parental involvement and the disadvantaged occupational position of women in society. If parents help daughters more, and female students are doing well academically, other stratifying mechanisms must be operating. It may be that college major is central to gender stratification. Jacobs (1995) has noted that gender differences in college majors have declined over time but not disappeared.
Future investigations in the area of parental involvement should aim at clarifying the motivation for the differential treatment of daughters and sons. Conducting in-depth interviews with parents might reveal whether their involvement with the education of daughters has been altered in reaction to social changes, or is a continuation of traditional socialization practices. This would contribute to a fuller understanding of gender differences in the educational experiences and outcomes of students.
The authors thank William C. Carter, Katharine M. Donato, Paul L. Leslie, Dawn T. Robinson, Charles M. Tolbert, II, and anonymous reviewers for their helpful comments on earlier drafts of this paper. This research was supported by a grant to the second author from the Spencer Foundation. The data, analysis, and views expressed are solely the responsibility of the authors.
Roger A. Wojtkiewicz, Department of Sociology, Louisiana State University, Baton Rouge, Louisiana.
Block, J. H. (1983). Differential premises arising from differential socialization of the sexes: Some conjectures. Child Development, 54, 1335-1354.
Bogenscbneider, K. (1997). Parental involvement in adolescent schooling: A proximal process with transcontextual validity. Journal of Marriage and the Family, 59, 718-733.
Catsambis, S. (1994). The path to math: Gender and racial-ethnic differences in mathematics participation from middle school to high school. Sociology of Education, 67, 199-215.
Cohen, J. (1987). Parents as educational models and definers. Journal of Marriage and the Family, 49, 339-351.
Downey, D. B., Jackson, P. B., & Powell, B. (1994). Sons versus daughters: Sex composition of children and maternal views on socialization. Sociological Quarterly, 35(1), 33-50.
Eccles, J. S., & Jacobs, J. E. (1986). Social forces shape math attitudes and performance. Journal of Women in Culture and Society, 11(2), 367-380.
Eccles, J. S., Jacobs, J. E., & Harold, R. D. (1990). Gender role stereotypes, expectancy effects, and parents' socialization of gender differences. Journal of Social Issues, 46, 183-201.
Entwisle, D. R., Alexander, K. L., & Olson, L. S. (1994). The gender gap in math: Its possible origins in neighborhood effects. American Sociological Review, 59, 822-838.
Epstein, J. L. (1991). Effects on student achievement of teachers' practices of parent involvement. In S. B. Silvern & B. A. Hutson (Eds.), Advances in reading/language research: A research annual. Literacy through family, community, and school interaction (Vol. 5, pp. 261-276). Greenwich, CT: JAI Press.
Fehrmann, P. G., Keith, T. Z., & Reimers, T. M. (1987). Home influence on school learning: Direct and indirect effects of parental involvement on high school grades. Journal of Educational Research, 80, 330-337.
Felson, R. B., & Zielinski, M. A. (1989). Children's self-esteem and parental support. Journal of Marriage and the Family, 51, 727-735.
Hanson, S. L. (1994). Lost talent: Unrealized educational aspirations and expectations among U.S. youths. Sociology of Education, 67, 159-183.
Harris, K. M., & Morgan, S. P. (1991). Fathers, sons, and daughters: Differential paternal involvement in parenting. Journal of Marriage and the Family, 53, 53 1-544.
Jacobs, J. A. (1995). Gender and academic specialties: Trends among recipients of college degrees in the 1980s. Sociology of Education, 68, 81-98.
Lorber, J. (1994). Paradoxes of gender. London: Yale University Press.
Mare, R. (1995). Changes in educational attainment and school enrollment. In R. Farley (Ed.), State of the Union: America in the 1990s, volume one--economic trends (pp. 155-213). New York: Russell Sage Foundation.
Marini, M. M., & Brinton, M. C. (1984). Sex typing in occupational socialization. In B. F. Reskin (Ed.), Sex segregation in the workplace: Trends, explanations, remedies (pp. 192-232). Washington, DC: National Academy Press.
Mickelson, R. A. (1989). Why does Jane read and write so well? The anomaly of women's achievement. Sociology of Education, 62, 47-63.
Morgan, S. P., Lye, D. N., & Condran, G. A. (1988). Sons, daughters, and the risk of marital disruption. American Journal of Sociology, 94(1), 110-129.
Muller, C. (1993). Parent involvement and academic achievement: An analysis of family resources available to the child. In B. Schneider & J. S. Coleman (Eds.), Parents, their children, and schools (pp. 77-113). Boulder, CO: Westview Press.
Muller, C. (1998). Gender differences in parental involvement and adolescents' mathematics achievement. Sociology of Education, 71, 336-356.
Muller, C., & Kerbow, D. (1993). Parent involvement in the home, school, and community. In B. Schneider & J. S. Coleman (Eds.), Parents, their children, and schools (pp. 13-42). Boulder, CO: Westview Press.
National Center for Education Statistics. (1994). Second follow-up: Student component data file user's manual (pp. H11-H12). Washington, DC: U.S. Department of Education, Office of Educational Research and Improvement.
Saltiel, J. (1985). A note on models and definers as sources of influence in the status attainment process: Male-female differences. Social Forces, 63, 1069-1075.
Schneider, B., & Coleman, J. (Eds.). (1993). Parents, their children, and schools. Boulder, CO: Westview Press.
Smith, T. E. (1992). Gender differences in the scientific achievement of adolescents: Effects of age and parental separation. Social Forces, 71, 469-484.
Snow, C. E., Barnes, W. S., Chandler, J., Goodman, I. F., & Hemptill, L. (1991). Unfulfilled expectations: Home and school influences on literacy. London: Harvard University Press.
Spain, D., & Bianchi, S. M. (1996). Balancing act: Motherhood, marriage and employment among American women. New York: Russell Sage Foundations.
Sui-Chu, E. H., & Willms, J. D. (1996). Effects of parental involvement on eighth-grade achievement. Sociology of Education, 69(2), 126--141.
Teachman, J. D., Paasch, K., & Carver, K. (1996). Social capital and dropping out of school early. Journal of Marriage and the Family, 58, 773--873.
Thomas, W. I., & Thomas, D. S. (1928). The child in America. New York: Alfred A. Knopf.
United States Department of Education. (1987). What works: Research about teaching and learning (2nd ed.). Washington, DC: Author.
United States Department of Education, National Center for Educational Statistics. (1996). Descriptive summary report, with an essay on access and choice in postsecondary education [National Education Longitudinal Study 1988-1994]. NCES 96--175, by A. Sanderson, B. Dugoni, K. Rasinski, & J. Taylor; D. C. Carroll, project officer. Washington, DC:
Useem, E. L. (1992). Middle schools and math groups: Parents' involvement in children's placement. Sociology of Education, 65, 263--279.
Wellesley College Center for Research on Women. (1992). The AAUW report: How schools shortchange girls --a study of major findings on girls and education. The AAUW Educational Foundation and National Education Association.
Wigfield, A., & Eccles, J. S. (1994). Children's competence beliefs, achievement values, and general self-esteem change across elementary and middle school. Journal of Early Adolescence, 14, 107--138.
Means for Parental Involvement Variables, Grades, Test Scores, and Educational Aspirations Unweighted Standard Weighted Mean Deviation Mean Dependent variables School discussion 4.292 1.082 4.284 Parent-School connection 1.532 1.011 1.499 Parental expectations 6.040 1.651 5.926 Parents check homework .734 .442 .740 Parents limit watching TV .393 .488 .372 Parents limit going out .727 .445 .728 Parental attendance .641 .480 .633 Key independent variables Grades 2.911 .753 2.887 Test scores 50.624 10.275 50.001 Educational aspirations 2.810 .972 2.746 Note: Means were calculated with missing data excluded. OLS Regressions of School Discussion, Parental Expectations, and Parent-School Connectivity School Discussion Parental Expectations b SE b SE MODEL 1 Male contrast contrast Female .170 [*] .014 .234 [*] .022 MODEL 2 Male contrast contrast Female .100 [*] .013 .067 [*] .019 Grades .188 [*] .011 -.019 .016 Test Scores .002 [*] .001 .012 [*] .001 Educ. Aspirations .220 [*] .008 .970 [*] .012 [R.sup.2] .150 .394 n 23,302 20,012 Parent-School Connectivity b SE MODEL 1 Male contrast Female -.102 [*] .014 MODEL 2 Male contrast Female -.109 [*] .014 Grades .008 .012 Test Scores -.006 [*] .001 Educ. Aspirations .082 [*] .009 [R.sup.2] .082 n 18,249 Note: Each model includes background variables and dummy variables and dummy variables for missing data. (*.)p [less than] .05 Logistic Regressions for Parental Supervision Variables and Parental Attendance Check Homework Limit TV Limit Going Out Coeff SE Coeff SE Coeff SE MODEL 1 Male contrast contrast contrast Female -.290 [*] .040 .015 .037 .221 [*] .039 MODEL 2 Male contrast contrast contrast Female -.285 [*] .041 -.021 .037 .195 [*] .040 Grades -.021 .034 .058 .031 .009 .033 Test Scores -.022 [*] .003 .008 [*] .002 .014 [*] .003 Educ. Aspirations .125 [*] .025 .113 [*] .023 .046 .024 -2 LL 14852.1 16873.3 15263.2 n 24,414 24,348 24,316 Parental Attendance Coeff SE MODEL 1 Male contrast Female .196 [*] .039 MODEL 2 Male contrast Female .081 [*] .040 Grades .443 [*] .033 Test Scores -.002 .003 Educ. Aspirations .284 [*] .024 -2 LL 14662.8 n 23,293 Each model includes background variables and dummy variables for missing data. (*.)p [less than] .05