How to prevent social desirability bias in surveys

Chances are that people tend to give socially desirable answers to direct survey questions on sensitive subjects, like watching porn, smoking during pregnancy, theft in the supermarket or illegal acquisition of prescription drugs. Pieters’ research concerned consumer acquisition of prescription drugs to enhance sexual performance through illegal channels such as internet sites. In earlier research, Pieters and his colleagues also studied pornography viewership. Consumers proved to be untruthful about the subject in direct questions. 

Truthful answers 

In order to get more truthful answers, the researchers developed a new questioning technique. They asked participants two questions: a ‘sensitive’ and an ‘innocent’ question. For example, in the study into medication to enhance sexual performance, these were: Question 1: Have you ever obtained medication via an illegal channel? Question 2: Have you ever bought products via the internet? 

Researcher Rik Pieters: ‘If you ask direct questions, you get two answers, for question 1 and question 2 separately. But some people may not give an honest answer to the first question. In our new, indirect technique, we only ask: ‘Is the answer to these two questions the same or different? So the respondents do not need to directly answer each of the two separate questions. That makes it easier for respondents to answer truthfully with respect to the first, sensitive question.’

Different questions

In addition, participants were also asked all kinds of different questions as part of the study, for example, on whether they liked particular activities or sports like football, baseball, chess, pole jumping across a stream (fierljeppen), or reading particular literature. The researchers used the answers to the other questions to predict whether, if respondents engage in illegal or sensitive activities, they would answer direct questions on these matters in the affirmative. This helped the researchers deduce what the likely answers to the sensitive questions would be.

The advantage of this technique is that survey participants answer sensitive questions truthfully, or at least more truthfully than under normal circumstances. Moreover, participants’ privacy is protected, and that is important with regard to sensitive or even illegal behaviors. Rik Pieters: ‘We never know exactly whether participants have engaged in the sensitive characteristic of interest, and that is not what we are interested in, either. But because of a survey conducted in this way, we do know how many people in the sample engaged in the sensitive behavior and with what other of their characteristics it correlates.’

Relevant information for policymakers

The technique can be applied in many areas, not only to study the illegal use of prescription drugs, but also to conduct research on money laundering, fraud, the purchase of illegal fireworks, drugs abuse, unhealthy behaviors, shoplifting, or harassment in the work environment. 

Pieters: ‘To be able to develop policy and assess its effects, public and private policymakers need to have reliable information on consumers’ preferences and behaviors. If survey respondents underreport or overreport certain behavior, policymakers do not have the information they need. This can lead to no policy being made (‘Nobody smokes in the Netherlands’) or the conclusion that a policy worked like charm (‘After the advertising campaign, nobody has reported having engaged in shoplifting’). That is why it is important that survey respondents answer sensitive questions truthfully.’

The paper ‘Response Aggregation to Obtain Truthful Answers to Sensitive Questions: Estimating the Prevalence of Illegal Purchases of Prescription Drugs’ was published in the leading academic Journal of Marketing Research. For more information and interview requests, please contact Rik Pieters (F.G.M.Pieters@tilburguniversity.edu).