ABSTRACT
In response to questions regarding methodology and sample posed by Hotz (this volume) and Upchurch and Kusunoki (this volume), this rejoinder provides more information on sample creation, sample selection, and the robustness of the results from our earlier chapter (Manlove, Ryan, & Franzetta, this volume). We also address issues related to unobserved heterogeneity. After conducting further analyses with extended samples, we find our results and conclusions to be robust. Our overall findings, regardless of sample, are that: (1) teens are inconsistent contraceptive users; (2) those teens who are consistent contraceptive users in one relationship may not be consistent in a subsequent relationship; and (3) relationship and partner factors are associated with contraceptive consistency.
