Varden E. Hadfield and Dr. Joseph A. Olsen, Sociology

Extensive cross-sectional research has examined marital status and family composition and their effects on academic performance. Many cross-sectional studies have found moderate correlations between family structure and both grades and test scores. A few studies have found when race and socioeconomic status are controlled, correlations remain for only certain family types. However, after a thorough library computer index search and Internet National Educational Longitudinal Study index search, ill studies located were cross-sectional and focused on specific aspects of family structure, without time order data to provide evidence for causation.

This study began with exploratory cross-sectional analysis of the 1988 National Educational Longitudinal Study (NELS) of 24,599 eighth graders and their parents, using a universe of all eight graders enrolled in the Spring of 1988. The NELS, sponsored by the U.S. Department of Education National Center for Education Statistics, used a two stage stratified sample to select schools and students, then surveyed the parents of these students. After missing cases were excluded, the baseline N was 22,433 eighth graders. research, the combined effects of several measures of family structure on a combined academic performance variable were analyzed. Results of one multiple regression analysis are summarized figure 1. Separate multiple regression analyses of grades and test scores showed similar patterns, except family structure had a slightly stronger effect on grades than on test scores. Analyses of missing values showed the effect of family structure on both grades and test scores would be stronger if missing values were included.

To measure the longitudinal effects of family structure changes on academic performance, I used longitudinal data from the 1992 NELS second follow-up survey-of the 1988 survey students, who were in twelfth grade in 1992. Their parents and schools were also surveyed in 1992. Several eighth graders had dropped out, did not answer questions, or were not available in 1992, so 60% of the original sample of eighth-graders provided data for family structure changes between 1988 and 1992. This left a sample size of 16,395 for most variables examined in this study. Initial cross-tabulation of various structure variables showed that students from single female parent families were disproportionately missing, and that married parent families were least likely to be missing. Also, students who have dropped out or who get lower grades are more likely to be missing. Therefore, the measured effects of structure changes are weaker than they would be if missing data could have been included in the analyses, especially for single-parent families.

In the longitudinal analysis, since many different hypotheses were involved, a multiple analysis of variance using effects contrast coding was used. Scheffe’s test for significant differences also was conducted. In these analyses, the effects of 1988, 1989, 1990, or 1991 family structure changes on 1992 grades, test scores, and socioeconomic status was examined.

In initial means comparisons, students whose parents stayed married had the best grades, test scores, and socioeconomic status. Surprisingly, the lowest grades were reported by students whose parents had remained “never married” during the entire four-year period. However, much of the reduced academic performance in these single-parent homes may be related to an extremely low socioeconomic status—about twice as low as widowed families, the second lowest group. Table 1 summarizes other significant findings of this comparison.

For students whose parents were divorced or separated, the difference in effect between grades and test scores is interesting. Test scores were affected most immediately after the divorce, while two or three years after the divorce, grades were most affected and test scores recovered. However, this difference between recent and less recent effects was not found for other family structure changes. Further research could explore the effects of recent verses earlier divorce on grades and test scores.

References

1. J.D. Finn and M.F. Owings, “Family Structure and School Performance in Eighth Grade,” Journal of Research and Development in Education,(March 1994)176-187.
2. S.A. Lee, “Family structure Effects on Student Outcomes,” IN Parents, Their Children, and Schools, Boulder, Colorado: Westview Press(1993)43-47
3. D.B. Downey, “Do Children in Single-Parent Households Fare Better Living with Same-Sex Parents?” Journal of Marriage and the Family,(February 1993)55-7 1.
4. T.Z. Keith, et al, “Does Parental Involvement Affect Eighth-Grade Student Achievement? Structural Analysis of National Data,” School Psychology Review, (1993)474-496.