Replies to comments made by Milne (see record 2007-00728-006) and Gray (see record 2007-00728-007) on the authors’ original article (see record 2007-00728-005). We value the critical appraisal of the findings in this article as part of the process of determining their value to clinicians. Our initial goal for this mixed-method analysis was to inform development of a forced choice, quantitative measure of predictors of adherence to preventive pelvic floor muscle training (PFMT) practice. This measure is now in use in a longitudinal follow-up study of intervention efficacy. We understand that some readers may be concerned that applying quantitative analysis to qualitative data violates some assumptions of inferential statistics, especially when there is missing data. For this reason, we present a conservative model where we considered all unknowns to be nonroutine and compared those in the routine category to those who could not be classified as routine. We hope that the combined data provide a basis for clinicians to make a solid recommendation for a routine approach in the interim, while waiting for confirmation from quantitative procedures where the amount of missing data is smaller. Conclusions from data collected from a forced-choice survey currently in use will have the advantage of fully satisfying the assumptions of inferential statistics. (PsycINFO Database Record (c) 2007 APA, all rights reserved)