Data for: Improving Abortion Underreporting in the United States: A Cognitive Interview Study

Résumé 0

Project Summary: The purpose of this study was to inform experimental testing of new approaches for measuring abortion in surveys in the United States, by improving our understanding of how women interpret and respond to survey items asking them to report their abortion history. Using cognitive interviews, we developed, tested, and evaluated various question wordings, as well as conducted card sort and vignette activities to further discern how participants’ understand and classify abortion. We aimed to test questions to clarify which experiences to report as an abortion; reduce the stigma and sensitivity of abortion; reduce the sense of intrusiveness of asking about abortion; or increase the motivation to report. Question wordings were newly developed or modified versions of existing survey questions. Results from the cognitive interviews were used to develop an experimental survey to further explore how to improve the accuracy of abortion reporting. Research reported in this publication was supported by the Eunice Kennedy Shriver National Institute Of Child Health & Human Development of the National Institutes of Health under Award Number R01HD084473. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Data Overview: We conducted cognitive interviews with 64 cisgender women in suburban Wisconsin (N=35) in January 2020 and urban New Jersey (N=29) in February 2020. We selected the two study states because of differences in abortion climate and to avoid geographically-specific findings. Participants were recruited to participate in an interview on sexual and reproductive health (SRH) by a third-party recruiting agency. Eligible participants were between the ages of 18 and 49, assigned female at birth, identified as women, spoke English, lived in Wisconsin or New Jersey, and had ever had penile-vaginal sex. Additionally, respondents were asked if they ever had an abortion during the screening process. The decision to include women with various abortion histories was deliberate. We sought to include respondents who may not have disclosed their abortion during the screening process and wanted feedback from all women of reproductive age, as that is the target sample of the experimental survey. We included feedback from both groups of respondents and highlighted their differences when applicable. The study was reviewed and approved by the Guttmacher Institute’s Institutional Review Board. Interviews lasted about 60-90 minutes and were conducted in private rooms at conference and market research locations by two members of the research team (AV and JM; see Interviewer Identities document). Interviews were audio recorded. We obtained verbal consent from all study participants, and they completed a short sociodemographic questionnaire at the end of the interview. Participants received $150 cash as a token of appreciation for their time. Sixteen respondents demonstrated accurate understanding of data sharing and consented to having their data be publicly shared; therefore, these are the only transcripts made available here. Data Collection Overview: The cognitive interview included several sections. In the first section, we asked participants to respond to and provide feedback on various versions of questions about their abortion history. In this section, they disclosed their abortion history directly to the interviewer. In the second section, we asked participants to provide feedback on several different introductory text options that might come before a question about their abortion history. In the third section, we asked participants more general questions about how they would prefer to be asked about whether they’d had an abortion on a survey, as well as their definition of abortion and why some people may choose not to disclose their abortion history. In sections 4 and 5, we conducted a card sort activity and went through a series of vignettes to further discern participants’ understanding and classification of abortion. Each interview was digitally audio recorded (.mp3). The audio recordings were uploaded from the devices to a folder on the Guttmacher network. The audio recordings will be uploaded to the secure, password-protected servers of the professional transcription agency, Datalyst LLC., for transcription. The audio recordings were deleted from the Datalyst servers upon completion of transcription and certification by Guttmacher staff that transcription was accurate/complete. Transcripts were cleaned by one of two team members; during this process, they also removed any potentially identifying information. Audio recordings were deleted from the devices at the end of data collection. The interviewer electronically wrote up immediate thoughts and impressions post-interview in memos (.docx) using a predetermined memo shell that mirrored the major sections of the interview guide. At the end of each interview, participants completed a demographic survey on paper. Data from the demographic surveys were entered into a combined demographic data spreadsheet (.xlsx). Organization of Shared Data: The following pieces of data were generated from all respondents: audio recording of interviewer, transcript of audio recording, interviewer memo of interview, demographic survey responses. An individual record is identified by the participant ID number, which is the link that connects all pieces of data. Participant ID numbers follow the naming convention of “State-Participant ID” (e.g., NJ-01). Data Analysis: Cleaned, de-identified transcripts, interviewer memos, and the demographic data spreadsheet were uploaded to an NVivo project (.nvp) hosted on the Guttmacher NVivo Plus network. To aid with coding of the data, we created a coding scheme (.xslx). From the analysis file, we generated node reports (.docx). The node reports were condensed into matrices (.xlsx). We reviewed node reports, commented on potential themes, and summarized potential themes at the end of each node report. We drafted an analytic memo for each node report that summarizes the emergent themes and wrote up the report with feedback on each question, particularly the approaches that seem most promising for future survey work.