Statewide Survey of
Women of Reproductive Age
in Delaware and Maryland

(“DELAWARE / MARYLAND SURVEY OF WOMEN”)

Questionnaire Development
Michel Boudreaux and Michael S. Rendall
University of Maryland, College Park
April 2019

Acknowledgments: This work was conducted in collaboration with NORC at the University of Chicago. It was supported by a research grant from a private philanthropic foundation.

Introduction

The Statewide Survey of Women of Reproductive Age in Delaware and Maryland was designed and conducted to support evaluation of the Delaware Contraceptive Access Now (DelCAN) initiative. The survey was administered to a probability sample of women aged 18 to 44 in Delaware and Maryland to measure the reproductive health experiences of women. The survey was fielded by NORC at the University of Chicago from November 2016 through March 2017. Eligible respondents were women aged 18-44 residing in households. The survey, branded as “Delaware / Maryland Survey of Women,” was predominantly fielded via the web with follow-up modes conducted via mail and telephone. Spanish and English versions of the survey were provided. All respondents self-reported. This report provides an overview of the questionnaire and its development. Information on the sample design, fielding operations, sample weighting, and other methodological characteristics of the survey are documented in a separate Methodological Report:
View / Download Baseline Survey Methodology Report

Goals of the Instrument

Few regularly administered surveys capture population representative information on contraceptive method use, beliefs, and reproductive health experiences. The statewide Survey of Women of reproductive age in Maryland and Delaware was created to obtain contraceptive method use and related data from a population based sample of adult women in Delaware and their counterparts in a neighboring state (Maryland) that were not exposed to the intervention. The survey was designed to gather information on current and lifetime contraceptive use; beliefs and attitudes about pregnancy prevention and pregnancy intention; sexual activity; pregnancy history; abortion attitudes and experiences; and socio-demographics.

Questionnaire Topics and Organization

The questionnaire was divided into 10 sections (see complete instrument in English or Spanish). The first section asked about age and sex and was used to screen for eligible respondents. The screener was followed, in order, by sections on General Health, Past Birth Control Use, Sources of Medical Information, Public Policy Opinions, Demographics, Reproductive Health, Current Birth Control Use, Past Pregnancies, and Prenatal Care.

Question Sources and the Construction of Selected Items

Whenever possible, question text and response items were obtained from existing surveys. Table 1 (below) provides an overview of sources used for a large set of selected items used in the survey. Basing question text and response items on existing surveys allows for estimates to be comparable with those from other surveys and takes advantage of the development and testing process used by those surveys. The construction of some of the key items is described below.

Birth Control Methods

The sections on past (“ever used”) and current birth control methods and reasons for not using birth control were adapted from the National Survey of Family Growth (NSFG). The section on past contraceptive use appeared before the current use questions. Each series asked about 13 distinct methods. A specific item measuring any use versus no use was provided in the current series: “Are you currently using any method or methods of birth control?” The question purposefully did not condition on use at last sex in order to reduce undercounting of permanent methods (Fabic & Becker, 2017). Respondents reporting yes were then asked to report the specific methods they use. Method types were provided using a laundry-list format that requested “Yes”, “No”, “Don’t Know” or “Prefer Not to Answer” responses to each method type. Withdrawal was purposefully placed first in the list to reduce false-negative reporting (Jones et al. 2009). The section on past use did not include a general any use question. Non-use in the ever section was inferred using the residual method (no use of any method type). Respondent’s recognition of specific method types was aided with pictures and detailed descriptions of birth control devices. In the paper copy of the instrument, the pictures and descriptions appeared at the start of the survey (prior to the screener). On the web version, the images of each method appear on the same web page that they are asked about the use of that particular method.

Because the DelCAN initiative includes a substantial emphasis on long-acting reversible contraceptives (LARC), the survey included follow-up questions on LARC. Brands of LARC used, reasons for non-use of LARC, and experiences of side-effects associated with LARCs were asked about. The current-method series included follow-up questions on method satisfaction. These items were original to the survey and tested as described in detail below.

Abortion List Experiment

Self-reports of abortion are known to be subject to substantial error (Lindberg & Scott, 2018). To overcome this error, a double list-experiment (Moseson et al. 2017) was used. The overall sample frame was randomly assigned to one of two versions of the survey instrument in which the respondents were presented with two lists. Each list contained a set of health care experiences and respondents were asked to report how many of the experiences applied to them. One list contained abortion and the other did not. The list of the health and abortion experiences were presented in different orders in the two versions of the survey. Responses across the two lists can be compared to generate an estimate of abortion prevalence, but the data cannot be used to infer the abortion experience of any given respondent.

Sources of Contraceptive Information

To facilitate evaluation of DelCAN’s public awareness campaign three questions were asked on about the sources and content of information respondents had received about birth control. One item (Question 27) asked about sources of information on birth control in the previous 3 months and included response options for 10 sources. A separate question (Question 28) asked about the kind of information received from these sources (e.g. where to obtain birth control, effectiveness, and cost). The third question (Question 29) asked if the respondent knew where to obtain free birth control methods. In the Delaware version of the survey only, these three questions were preceded by a question (Question 26) asking respondents if they been to any of a comprehensive list of Delaware Title X funded clinics.

Translation

All questions were originally written in English. The instrument was translated into Spanish by a an organization that NORC contracts with to translate their materials. Then, approximately four individual translators reviewed and edited the original translation. As described below, both the English and Spanish versions underwent cognitive testing. During that phase additional edits were made to the translation.

Question Testing and Cognitive Interviews

After an initial draft of the questionnaire was constructed it was informally fielded to colleagues and students at the University of Maryland. Refinements were made and then a subsequent version underwent cognitive testing by NORC at the University of Chicago with women of reproductive age living in and around the Chicago area to test for reader comprehension and clarity of the survey items. Eleven subjects age 18-44 residing in the Chicago area were recruited from Craigslist. Interviews, using a “think-aloud” method, were conducted using the paper questionnaire. Eight interviews were based on the English survey and three on the Spanish survey. Cognitive interviewing led to specific revisions to the questionnaire such as including the phrase “tubes tied” in the item about tubal sterilization, unbundling of double-barreled questions, and simplifying a number of items. In addition, the inclusion of pictures representing each birth control were added to the questionnaire after identifying that many women, even those using contraception, often cannot correctly name the type of method they are using (e.g., implant versus IUD).

View / Download the Survey of Women (English)

View / Download the Survey of Women (Spanish)

Source of Questions table is given below references

View / Download a copy of this Introduction page

References

Moseson, H. Treleaven, E. Gerdts, C. Diamond-Smith, N. 2017. The List Experiment for Measuring Abortion: What We Know and What We Need.” Studies in Family Planning 48(4): 397-405.

Lindberg, L. Scott, R. 2018. Effects of ACASI on Reporting of Abortion and Other Pregnancy Outcomes in the US National Survey of Family Growth.” Studies in Family Planning 49(3): 259-278.

Fabic, M. Becker, S. 2017. “Measuring Contraceptive Prevlance among Women Who are at Risk of Pregnancy.” Contraception 96:183-188.

Jones, R. K., Fennell, J., Higgins, J. A., & Blanchard, K. (2009). Better than nothing or savvy risk-reduction practice? The importance of withdrawal. Contraception, 79(6), 407-410.

Middlestadt, S. E., Bhattacharyya, K., Rosenbaum, J., Fishbein, M., & Shepherd, M. (1996). The use of theory based semi-structured elicitation questionnaires: Formative research for CDC’s prevention marketing initiative. Public Health Reports, 111(Suppl 1), 18.

Page last updated: May 29, 2020