Sexual activity and contraceptive use among low-income urban black adolescent females
Keith, J.B.; McCreary, C.; Collins, K.; Smith, C.P.; Bernstein, I.
Adolescence 26(104): 769-785
1991
ISSN/ISBN: 0001-8449 PMID: 1789165 Document Number: 320379
The findings of this longitudinal study of contracepting behavior and attitudes among black and Hispanic females are based on an initial data sample. Study respondents (142 black never-pregnant youths 13-17 years) were recruited from the West Dallas Youth Clinic, a well-established school-based comprehensive health service serving a low income population. Interviews were conducted using the Nathanson and Becker's (1983) HBM questionnaire, Nowicki and Strickland's Locus of Control measure, a Piagetian Logical Reasoning Test, and Stein el al.'s (1968) Future Events Test. 3 groups were differentiated: nonsexually active (NA) (50), sexually active and noncontracepting (A/NC) (20), or sexually active and contracepting (A/C) (72). Initial analyses revealed age differences among the groups (F (2,139) =14.51, p.01). LInear and logistic regression analyses determined in general that group differences could not be attributed to age differences. The mean was 15.1 years for NA, 15.7 years for A/NC, and 16.1 years for A/C. Groups did not differ significantly academically on, e.g., school enrollment, grades, and outside job variables. Maternal characteristics were unrelated to group differences, but having a father living at home was, i.e., 23.8% of A/C had a father living at home vs. 45.8% of N/As and 38.1% of A/NCs. Because motivation scales were unreliable, future research and development are necessary. Nonetheless scale means varied among the groups, i.e., where N/As had higher motivation than A/NCs, and with A/Cs similar to both groups. The Perceived Probability of Pregnancy scale was very heterogenous. A/Cs felt a greater threat of pregnancy and were more likely to practice contraception than A/NCs. In the intercorrelations analysis, positive attitudes were related to a lower probability of sexual activity (r=.36). Those with a more positive attitude toward contraception were more sexually active. A boyfriend as a source of information was positively correlated with reported sexual activity (r=.38). Contraceptive use was associated with the Perceived Probability scale (r=.40), the Attitudes scale (r=.20), the Total Benefits and Barriers scale (r=.35), age (r=.26), and being sexually active (r=.38). Path analysis was deemed inappropriate. Sexual activity was predicted in multiple correlation analyses using separate Benefits and Barriers scales (R=.68, F (12,129) =9.23, p.001). Contraceptive use was also significantly predicted but with significant beta weights only for the General Preference subscale. The General Preference subscale, age, and boyfriend as information source were positively associated with sexual activity, and family influences decisions were inhibitory. The presence of a natural father had nonsignificant beta weight, and is interpreted as the father's role is indirect in inhibiting sexual activity.