Participants were 582 mother-infant dyads from the Par-enting for the First Time Project, a multi-site, longitudinalstudy following primiparous mothers and children from theprenatal period through the third year of life. Mothers wererecruited from hospitals, health clinics, social service agen-cies, and school-aged mothers programs in South Bend, IN,Washington, D.C., Kansas City, KS, and Birmingham, AL.The sample consisted of 338 adolescent mothers and 244adult mothers with a wide range of educational back-grounds. Mothers ranged in age from 15 to 35 years (meanage
21.28) at time of childbirth and were ethnicallydiverse (65% African-American, 16% Caucasian, 15%Latina, and 4% Multi-Ethnic). During the prenatal period,the majority of mothers were single (63%), although manymothers were living with their partner (21%) or married(16%). In terms of educational attainment, 44% of mothershad not completed high school, 25% had a high schooldiploma or equivalency, and 31% had some college orvocational education. Similarly, 30% of fathers had notcompleted high school, 41% had a high school diploma orequivalency, and 28% had some college or vocational edu-cation. Fifty-six percent of mothers were living with theirown mothers when children were 4 months old. Approxi-mately 70% of families reported annual incomes below$20,000 (37% below $10,000) and only 13% earned morethan $40,000 annually when children were 6 months old.Participants were included in analyses if they had datafrom any one of the time points of interest. Full InformationMaximum Likelihood Estimation was utilized to addressmissing data. The 4-, 6-, 8-, 12-, 18-, and 24-month assess-ments had 18%, 27%, 27%, 24%, 29%, and 37% missingdata, respectively, on the father support variable. This levelof missing data is comparable to other longitudinal studiesof high-risk populations (e.g., Hansen, Tobler, & Graham,1990; Kilpatrick, Acierno, Resnick, Saunders, & Best,1997). Participants with missing data did not differ fromthose with complete data in terms of annual family income,
.05. However, mothers with completedata were older,
.05, and had highereducational attainment,
.01, thanmothers with missing data. Subsequently, maternal age andeducational attainment were included as covariates in thelongitudinal analyses.
Design and Procedure
The present study utilized data gathered from mothersduring their third trimester of pregnancy and when childrenwere 4, 6, 8, 12, 18, and 24 months of age. Multiple effortswere taken to protect the rights of participants in accordancewith IRB requirements. Consent forms were signed at theprenatal assessment, and mothers were informed that theyhad the right to refuse any part or all of the subsequentassessments. For mothers who were younger than 18 yearsof age, parental consent forms were signed in addition toparticipant assent forms. Consent forms were resigned ateach additional interview.During the prenatal assessment, mothers were inter-viewed in the university laboratory. For assessments occur-ring when children were 4, 8, and 18 months of age, motherswere interviewed in their homes. When children were 6, 12,and 24 months of age, mothers and their children wereinvited to come to the university setting. Interviews typi-cally lasted up to two hours; transportation was provided forparticipants when requested. Although the majority of laboratory-based assessments occurred in the university set-ting, in cases where the mother was unable to come to theuniversity, the interview occurred in her home.For the initial interview during the third trimester of pregnancy, each mother provided basic demographic infor-mation about herself and the father of her child, includingage, education level, and father residential status. As part of a larger assessment protocol at each time point (4, 6, 8, 12,18, and 24 months), mothers reported on the amount of support they received from their baby’s father. In addition,at the 6-, 12-, and 24-month measurement points, mothersresponded to self-report items measuring depressive symp-tomatology. At every interview, families were compensatedfor their participation with Wal-Mart gift cards. Participantswere also contacted via telephone between interviews tomaintain rapport and reduce attrition.
Social support from father.
Social support from fatherswas assessed utilizing 6 items drawn from the Life HistoryInterview, an informal interview developed for the largerstudy. For each item, mothers indicated with a yes (coded as1) or no (coded as 0) response whether the father of theirbaby provided support. Items were as follows:Does child’s father provide ﬁnancial or part-ﬁnancialsupport?Does child’s father provide diapers, gifts, food, etc?Does child’s father provide help with childcare on aregular basis?Does child’s father visit the child?Does child’s father provide help with transportation?Does child’s father provide by his family helping takecare of the baby?The dichotomous responses for each item were summed,creating a possible range of scores from 0 to 6, with higherscores indicating greater levels of support from fathers.Cronbach’s alphas for the 6 items ranged from .87 to .89across the 6 time points.
The BDI-II (Beck, Steer, & Brown, 1996)is a new edition of the widely used Beck Depression Inven-tory (Beck & Steer, 1984) and has items that relate todepression criteria of the
Diagnostic and Statistical Manual