
Funded by the German Research Foundation (WE 5757/4-1, DFG KA 5005-1/1)
About this Research Project
The aim of this study was to investigate how small differences in women’s voices related to pregnancy can be detected by others. Therefore, Swedish and German female speakers had been recorded both during their pregnancy and six months after birth. During both recording sessions, speakers were asked to describe the same set of pictures containing certain target words. For each speaker, we choose utterances including the same three target words (German: <Tiger>, <Kuh>, <Tasse>; Swedish: <fisk>, <ko>, <katt>) for the first and the second recording session and created three utterance pairs with each pair containing the same target word. In the perception experiment, female raters were asked to judge each utterance pair regarding one of four assessment dimensions (pregnancy, femininity, competence, fatigue). Raters were randomly assigned to one assessment dimension. The order in which the utterance pairs were presented to each rater varied.
We expect acoustic-phonetic parameters to differ between the pre- and postpartal recordings which would then be used by the raters to accurately recognize women’s pregnancy based on their voices. Furthermore, we assume that pregnancy status could also impact the assessment of the speaker’s femininity, fatigue, and competence.