School of Economics, University of Bristol, Bristol BS8 1TU, United Kingdom.
Erasmus School of Economics, Erasmus University Rotterdam, Rotterdam 3062, Netherlands.
Proc Natl Acad Sci U S A. 2024 Aug 20;121(34):e2407629121. doi: 10.1073/pnas.2407629121. Epub 2024 Aug 13.
Desired fertility measures are routinely collected and used by researchers and policy makers, but their self-reported nature raises the possibility of reporting bias. In this paper, we test for the presence of such bias by comparing responses to direct survey questions with indirect questions offering a varying, randomized, degree of confidentiality to respondents in a socioeconomically diverse sample of Nigerian women ([Formula: see text]). We find that women report higher fertility preferences when asked indirectly, but only when their responses afford them complete confidentiality, not when their responses are simply blind to the enumerator. Our results suggest that there may be fewer unintended pregnancies than currently thought and that the effectiveness of family planning policy targeting may be weakened by the bias we uncover. We conclude with suggestions for future work on how to mitigate reporting bias.
研究人员和政策制定者通常会收集和使用期望生育率数据，但由于这些数据是自我报告的，因此可能存在报告偏差。在本文中，我们通过比较直接调查问题和间接问题的回答，在一个社会经济多样化的尼日利亚女性样本中，为受访者提供不同程度的、随机的、保密程度的间接问题，以此来检验是否存在这种偏差([公式：见正文])。我们发现，当被间接询问时，女性报告的生育意愿更高，但只有当她们的回答完全保密时才会这样，而不是简单地对调查员隐瞒时才会这样。我们的结果表明，可能存在的意外怀孕比目前认为的要少，而我们发现的这种偏差可能会削弱以目标人群为对象的计划生育政策的有效性。最后，我们提出了关于如何减轻报告偏差的未来工作建议。
