Statistics in Brief:

The Relationship Between the Parental and Marital Experiences of 1988 Eighth-Grade Girls and High School Completion as of 1994

May 1998

(NCES 98-096) Ordering information


The relationship between teenage childbearing and young women's educational and occupational attainment has been well established. In general, young women who give birth as teenagers are less likely to complete high school than women who delay childbearing until their 20s (Mott and Marsiglio 1985), regardless of a teenage mother's marital status (Teti and Lamb 1989). This Statistics in Brief describes the parental and marital experiences of the female members of the 1988 eighth-grade cohort (those females who were eighth-graders in 1998) as of 1994; it also describes their high school completion rates, as well as the timing of first births and first marriages in relation to June 1992, when most of this cohort should have completed high school. The National Education Longitudinal Study of 1988 (NELS:88), a nationally representative sample of 1988 eighth-graders who were followed up 2 years later in 1990 and again in 1992 and 1994, provides an opportunity to consider how motherhood and marriage were related to the educational attainment of the 1988 eighth-grade females. Findings include the following:

Overall Rates of Childbearing and Marriage of 1988 Eighth-Graders by 1994

Table 1 displays the parental and marital experiences of female members of the eighth-grade class of 1988. By 1994 (when the average age of this cohort was 20 years old), 22 percent of all 1988 eighth-grade females had given birth to at least one child, and 14 percent had been married at least once. Furthermore, 11 percent of all 1988 eighth-grade females had their first child before June 1992, when this cohort of eighth-graders was scheduled to complete high school. In contrast, 5 percent had been married at least once by June of 1992. About 7 percent had children before June of 1992 and had not married by 1994, while less than 1 percent married before June 1992 and not had any children before 1994. About 3 percent married and had children before June of 1992. Some 38 percent of females who had a child by 1994 were also married by 1994. (38 percent or 8.3/21.8 in table 1).

One advantage of NELS:88 over previous longitudinal studies are the relatively large samples of Asians and Hispanics. Past research on this topic has generally been confined to racial-ethnic comparisons between blacks and whites; however, interesting patterns among racial-ethnic groups emerged from these data. The samples of non-white racial-ethnic groups are still too small to give reliable estimates for combined marriage and childbearing separately for each group. Therefore, the following analysis examines the characteristics of marriage and childbearing separately for these groups. For example, table 2 displays the parental and marital experiences of female members of the eighth-grade class of 1988 by various demographic characteristics. Asian eighth-grade females were least likely among all racial-ethnic groups to have had a child by 1994, with 9 percent having had at least one child (table 2). Moreover, white eighth-grade females were less likely than their Hispanic and black counterparts to have had a child (17 percent compared with 34 and 39 percent, respectively).

Examining first marriages, however, reveals a different pattern. By 1994, 7 percent or fewer of both Asian and black 1988 eighth-grade females had been married at least once, compared with 20 percent of Hispanics and 15 percent of whites.

The likelihood of 1988 eighth-grade females having at least one child or marrying at least once by 1994 was also related to family socioeconomic status (SES). Ninety-three percent of females from families with high SES had not given birth to a child by 1994, compared with 81 percent from middle SES families and 63 percent from low SES families. Likewise, 94 percent of females from high SES backgrounds had never married, compared with 85 percent of those with middle SES and 79 percent with low SES.

Table 2 also reveals differences in childbearing and marriage by urbanicity and region. Females from rural areas were more likely to have had a child than suburban, and rural females were more likely than others to have been married at least once. Finally, females from the Northeast were less likely to have had a child or to have been married by 1994 than those who were from any other region while southern 1988 eighth-grade females were most likely to have been married.

Timing of Childbearing and Marriage

While overall rates of childbearing and marriage are worth noting, early childbearing and marriage are perhaps of more interest when examining high school completion. Figures 1 and 2 show the timing of parenthood and marriage relative to on-time progression through high school. Table 2 also displays the proportions of females who had given birth to their first child or had married for the first time before or after June 1992, when this cohort was expected to complete high school (also represented by the vertical lines in figures 1 and 2). Hispanic and black females were more likely to have had at least one child before June of 1992 than whites and Asian females. The same pattern held for females from low SES families compared with their middle and high SES counterparts (table 2). Moreover, 9 percent of females from low SES families had married for the first time before June 1992. Also, females from rural areas, as well as females from the South, were more likely to have married before the time they were expected to complete high school.

High School Completion Rates

As stated above, previous research has shown that teenagers, regardless of marital status, who have children are less likely to complete high school than those who postpone childbearing until adulthood. The data in table 3 reveal that the same was true for the female members of the eighth-grade cohort of 1988. Among all those who had given birth to their first child before June 1992, 54 percent had completed high school as of 1994. Females who had a child after June 1992 were more likely to have completed high school (74 percent). The same pattern held for first marriages: 48 percent of females who had married before June 1992 had completed high school compared with 83 percent who had married after high school. However, most of this difference reflects the fact that 65 percent of those who married before June 1992 also had a child before June 1992 (65 percent or 3.0/4.6 in table 1) and 40 percent of these females who had married and had a birth before June 1992 had completed high school by 1994. However, this percentage is not statistically different than the rate for those who had a child but did not marry before June 1992, perhaps reflecting the relatively small sample size of such females.

In contrast, 95 percent of females who were neither married or had children by 1994 had completed high school (by earning a diploma or an alternative credential). Having a child before scheduled graduation was much more detrimental to eventual completion than getting married before graduation-58 percent of females who had a child before June 1992 but never married had completed by 1994, compared with 84 percent who had married before scheduled graduation but did not have children. Also, for those females who had a child by 1994, there was not a significant difference between females who did not marry and those that did before 1994. (66.1 percent vs 60.3)

While in table 4 there were no differences in the proportions of eighth-grade females who had completed high school as of 1994 by urbanicity and only a small difference between females in the South compared with females in the Northeast, the proportions did vary by race-ethnicity. For example, among all 1988 eighth-grade females, Hispanics were less likely than Asians and whites to have completed high school by 1994 (78 percent compared with 93 and 90 percent, respectively). Blacks were also less likely to have completed high school than whites (84 percent compared with 90 percent). While this pattern remained for those who did not marry as of 1994, racial-ethnic differences in high school completion rates disappeared among those who did not have any children as of 1994.

There was also a direct relationship between high school completion rates and SES for all 1988 eighth-grade females. Although this relationship remained for those with no children, the differences were reduced.

The high school completion rates for whites and blacks who had a child before June 1992 did not differ significantly. But other racial-ethnic and socioeconomic variation in completion rates persisted among those who had given birth to their first child before June 1992. For example, for those who had a child by June 1992, 37 percent of Hispanic females completed high school, compared with 57 percent of white and 63 percent of black females. While two-thirds of females from middle SES families completed high school by 1994, 44 percent of females from low SES families did so. Finally, the high school completion rates among Hispanics, blacks, and whites who had married at least once by 1994 did not differ significantly, regardless of when they had married.

Conclusion

This report examined the relationship between the family formation patterns of females in the 1988 eighth-grade class and their high school completion rates as of 1994. The analysis revealed that the majority of the female members of this cohort had not begun to form their own families either by having a child or getting married by 1994. However, the racial-ethnic analysis revealed significant differences among groups. In particular, the results show that black and Hispanic females are more likely than their white and Asian counterparts to begin their families at young ages. More specifically, 39 percent of black females and 34 percent of Hispanic females had given birth to their first child by 1994, a time when the majority of this cohort was approximately 20 years old. Black and Hispanic females were also more likely to have had their first child before June 1992, when the majority of this cohort was scheduled to complete high school. Given past research showing the strong association between early family formation, female-headed households, and poverty, these results may suggest later disadvantages for this portion of black and Hispanic female members of the 1988 eighth-grade class.

The analysis of the relationship between timing of first births and first marriages and high school completion also revealed racial-ethnic patterns. While racial-ethnic differences in high school completion rates are well known and were confirmed in the analysis of all 1988 eighth-grade females, the differences disappeared among those who had not had any children as of 1994. Black and Hispanic females who had not had any children as of 1994 were just as likely to complete high school as white and Asian females. However, among those females who had given birth to a child before June 1992, 37 percent of Hispanics and 44 percent of those from low socioeconomic backgrounds completed high school. Future research should examine whether these students became pregnant after they dropped out or whether they dropped out following their pregnancies. Furthermore, the results of this report confirm past research showing that females who postpone childbearing and marriage until they reach their early 20s are more likely to attain their high school diploma.

References

Ingels, S. et al. NELS:88 Second Follow-Up Student Component Data File User's Manual. (1994). NCES 94-374.

Mott, F. and Marsiglio, W. "Early Childbearing and Completion of High School." Family Planning Perspectives, 17 (5) (1985): 234-237.

Teti, D. and Lamb, M. "Socioeconomic and Marital Outcomes of Adolescent Marriage, Adolescent Childbirth, and their Co-Occurrence." Journal of Marriage and the Family 51 (1989): 203-212.

Appendix: Technical Notes for NELS:88

The purpose of the National Education Longitudinal Study of 1988 is to collect trend data on students, dropouts, parents, teachers, and schools. The 1988 sample consisted of more than 24,000 eighth-grade students and more than 1,000 eighth-grade schools. For a detailed description of NELS:88, see Ingels et al. (1994).

Analysis sample

The results of this report can be generalized to the female members of the eighth-grade cohort of 1988. These students were identified with the variable F3SEX, and the analysis was performed using a final unweighted sample of 7,286. Of this total, 1,012 (312 prior to June 1992) had given birth and 1,399 (683 prior to June 1992) had been married by the spring of 1994. In addition, 600 (221 prior to June 1992) had both given birth and were married by the spring of 1994. Although included in the totals, American Indians/Alaskan Natives were excluded from the racial-ethnic analysis because the sample size was too small for reliable estimates. Data were weighted with F3QWTG8.

Variables Used in the Analysis

The student background variables used in this analysis include F3RACE (student racial-ethnic identification); G8URBN (urbanicity); G8REGON (region); and BYSESQ (family SES). To produce three categories of SES, the two middle quartiles of BYSESQ were collapsed into one category to be identified as "Middle SES."

The variable F3CHLLDT was used to identify those students who had given birth to at least one child. This derived variable contains the date of birth for the respondent's first child. It also identifies those who had at least one child, but the child's birthdate was unknown. Finally, it identifies those who did not have any children. A respondent was considered to have had at least one child if she had a valid birthdate for this child, or if she was identified as having a child, but the birthdate was missing. To understand first marriage patterns, the variable F3MARDT was used. For the purpose of this report, respondents were considered to have been married at least once if they had a valid marriage date, or if they were identified as having been married, but the date of marriage was missing.

High school completion status was measured with the variable F3DIPLOM. For the purpose of this report, "Received GED" and "Received Certificate" were combined with "Received a Diploma" into one category to describe those who had completed high school as of 1994.

Statistical procedures

All comparisons discussed in the text of this report have been tested for statistical significance to ensure that the differences did not occur by chance. For all multiple comparisons, the Bonferroni procedure was used.

The SUDAAN program was used to calculate all standard errors in order to adjust for the effects of the complex survey design of NELS:88.

Acknowledgments

The authors wish to express their gratitude to the various reviewers of this report. In particular, a special thanks to Ellen Bradburn and Mary Frase who provided many valuable suggestions and comments before and throughout the adjudication process. We would also like to thank the following individuals who served as the principal reviewers and who also provided valuable criticisms and helpful suggestions: Robert Burton of the Statistical Standards and Methodology Group; Michael Ross of the Surveys and Cooperative Systems Group; Edith McArthur of the Data Development and Longitudinal Studies Group; Jennifer Manlove of Child Trends Inc.; and Kathleen White of the GAO Education and Employment Division.


Table 1.-Percentage of 1988 eighth-grade females who had given birth to a child or had been married at least once as of 1994, by timing of event

Table 2.-Percentage of 1988 eighth-grade females who had given birth to a child or had been married at least once as of 1994, by timing of event and various background characteristics

Table 3.-High school completion rates as of 1994 for 1988 eighth-grade females by childbearing and marital status, by timing of event

Table 4.-High school completion rates as of 1994 for 1988 eighth-grade females by childbearing and marital status, by timing of event and various background characteristics

Figure 1.-Cumulative percentage of 1988 eighth-grade females who had given birth to a child by June 1994, by date of first birth, by race-ethnicity

Figure 2.-Cumulative percentage of 1988 eighth-grade females who had been married by June 1994, by date of first marriage, by race-ethnicity

Table A.-Standard errors for table 1-Percentage of 1988 eighth-grade females who had given birth to a child or had been married at least once as of 1994, by timing of event

Table B.-Standard errors for table 2-Percentage of 1988 eighth-grade females who had given birth to a child or had been married at least once as of 1994, by timing of event and various background characteristics

Table C.-Standard errors for table 3-High school completion rates as of 1994 for 1988 eighth-grade females by childbearing and marital status, by timing of event

Table D.-Standard errors for table 4- High school completion rates as of 1994 for 1988 eighth-grade females by childbearing and marital status, by timing of event and various background characteristics


For more information about the content of this report, contact Jeffrey Owings at Jeffrey.Owings@ed.gov.