Correlates and Consequences of Parent–Teen Incongruence in Reports of Teens' Sexual Experience

Abstract

Using the National Longitudinal Study of Adolescent Health, factors associated with incongruence between parents' and adolescents' reports of teens' sexual experience were investigated, and the consequences of inaccurate parental knowledge for adolescents' subsequent sexual behaviors were explored. Most parents of virgins accurately reported teens' lack of experience, but most parents of teens who had had sex provided inaccurate reports. Binary logistic regression analyses showed that many adolescent-, parent-, and family-level factors predicted the accuracy of parents' reports. Parents' accurate knowledge of their teens' sexual experience was not found to be consistently beneficial for teens' subsequent sexual outcomes. Rather, parents' expectations about teens' sexual experience created a self-fulfilling prophecy, with teens' subsequent sexual outcomes conforming to parents' expectations. These findings suggest that research on parent–teen communication about sex needs to consider the expectations being expressed, as well as the information being exchanged.

This research uses data from Add Health, a program project designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris, and funded by Grant P01–HD31921 from the National Institute of Child Health and Human Development, with cooperative funding from 17 other agencies. Special acknowledgment is due to Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. Persons interested in obtaining data files from Add Health should contact Add Health, Carolina Population Center, 123 W. Franklin St., Chapel Hill, NC 27516-2524 (www.cpc.unc.edu/addhealth/contract.html). We thank Richard Rogers and Jeff Dennis for their helpful comments.

Notes

¹It is unfortunate that the survey defined “sexual intercourse” more specifically for teens but not for parents. Furthermore, the survey's definition of sexual intercourse excluded all homosexual acts, as well as anal and oral heterosexual acts that many people include in their definitions of sexual intercourse (Bogart, Cecil, Wagstaff, Pinkerton, & Abramson, 2000; Sanders & Reinisch, 1999). Add Health did not include questions about oral sex in either wave, and anal sex experience was only recorded at Wave II. Supplementary analyses using Wave II teen reports and Wave I parent reports (the only wave available) suggested that including both anal and vaginal sex provoked little change in levels of congruence between parent and teen reports compared to vaginal sex only.

²We compared multivariate models presented in Tables 3 and 4 and Figure 2 with models using analysis samples that listwise deleted all missing cases instead of including missing data indicators on some variables. The only exception is that we still include a missing data indicator for grade point average (GPA) because that missing data indicator includes many valid cases in which students cannot report a GPA (e.g., if their school does not give traditional grades). For the main results testing hypotheses, the sign and significance does not differ across the two samples except that in the analysis of the effect of underestimation on the likelihood of having sex with drugs equivalent to Table 4, the effect of parents having control over decisions about the teens' life was no longer significant, but had the same sign and similar magnitude.

Note. Source: National Longitudinal Study of Adolescent Health (1995–1996; Bearman, Jones, & Udry, 1997).

Weighted means account for sample design effects (stratification and clustering).

The difference in accuracy of parental reports by level of Wave I sexual experience is significant (p < .001).

*p < .05. **p < .01. ***p < .001, design-based F test, within-row comparisons.

Note. Source: National Longitudinal Study of Adolescent Health (1995–1996; Bearman, Jones, & Udry, 1997).

Analyses account for sample design effects (weighting, stratification, and clustering).

^a n = 7,228. ^b n = 3,116.

*p < .05. **p < .01. ***p < .001; two-tailed tests.

Note. Source: National Longitudinal Study of Adolescent Health (1995–1996; Bearman, Jones, & Udry, 1997).

Coefficients, not odds ratios, are reported. Analyses account for sample design effects (weighting, stratification, and clustering). All dependent variables but the first are restricted to those who had sex between waves; the last is further restricted to females. Models control for teen age, gender, race or ethnicity, family structure, religious attendance, relationship status, grade point average, relationship with parent, and parent employment and education.

*p < .05, **p < 0.01, ***p < .001; two-tailed tests.

³Odds and percentages reported here and again later were calculated using odds ratios by exponentiating the logistic regression coefficients from Table 4 or Figure 2 and subtracting from 1 when applicable (in this case, e ^−0.2 = 0.82 and 1 − 0.82 = 0.18).

⁴Although this last result is statistically significant at the p < .01 level, it is based on 31 overestimating parents. Therefore, these results should be treated as preliminary.

⁵Supplemental analyses not shown restructured the congruency categories such that if parents were wrong by more than one year when predicting the age at which their child first had sex, they were moved into the incongruent category. This restructuring did not change our findings, nor did it largely change the size of the categories.

⁶Analyses not shown examined the effects of teens' beliefs that their peers would respect them more if they had sex (a proxy for peers' social norms) and of teens' perception of peers' familiarity with different contraception methods (a proxy for peers' behavioral norms) and found that these variables did not alter the relationships between overestimation or underestimation and sexual outcomes.