Pode a participação em estudos longitudinais de coortes alterar o comportamento dos participantes? Evidências para gravidez na adolescência com a pesquisa da coorte de Pelotas de 1982

Abstract

This paper investigates the effect of participation of female adolescents, aged 15, in a survey with questions related to sexual behavior on the probability of having had children from age 16 to 19 and to get pregnant up to 22 years of age. The participants in a population-based prospective epidemiological study, the 1982 Pelotas cohort, are analyzed. The sample consists of 1,945 adolescents. Given the occurrence of an unintended experiment, observed in the 6th follow-up of the 1982 cohort (which occurred in 1997), the empirical strategies employed were probit model estimation and propensity score matching (PSM). In both methods applied, the results showed that participation in the 6th follow-up, in which questions about sexual behavior were asked, did not change the likelihood of the participants having had children in adolescence, between the ages of 16 and 19, or to get pregnant up to 22 years of age.