PROJECTS

Improving Survey Quality in Qatar: Evaluating interviewers’ role in establishing interview privacy and its effect on survey response

Surveys that are designed to collect sensitive information or to measure the respondent’s own knowledge usually call for a private interview setting. Privacy is an important feature of the interview due to its effect on reporting information, especially sensitive information. Respondents interviewed in the presence of a third person could alter their responses leading to measurement bias. Moreover, the variation in interview privacy within a sample changes the experience across respondents, leading to differences in the measurement properties of the collected information. Researchers rely heavily on interviewers to ensure a private interview setting. In reality, interviewers are essentially guests in respondents’ homes and might find it difficult to ensure privacy. Achieving privacy usually requires some negotiation between the interviewer and household members. Interviewers who have a certain set of social skills, who are friendly negotiators, and who seem trustworthy might be better at achieving privacy. Interviewers’ traits, skills, and attitudes towards privacy, however, could vary leading to large between-interviewer variation in interview privacy, increasing the measurement error in the data collected. In spite of its impact on data quality, we know very little about interview privacy in Qatar and the ways in which it influences reporting. This becomes even more important in a society where people are greatly integrated into strong, cohesive in-group circles, such as family circles. In such societies, privacy among in-group members is usually less common and conformity to social norms is expected, leading to possible data biases. The issue of interview privacy and social conformity might be compounded in a survey environment that is dominated by a mismatch between the ethnicity of the interviewer and the respondent. Therefore, in a society that is witnessing an exponential reliance on scientific data to inform policies in many fields including health and social sciences, understanding such biases and reducing their impact to improve data quality becomes highly critical.

The SRC IU is supporting efforts to investigate the following: 1) the rate of non-private interviews in a nationally representative survey in Qatar, 2) respondent-level and interviewer-level predictors of interview privacy, 3) the effect of interview privacy on reporting sensitive information. Most importantly and based on the research findings, this proposal aims at devising culturally-relevant interviewer training material that will give interviewers the knowledge and skills needed to ensure privacy leading to better data quality.