Increasing accurate responses to sensitive research questions

Can creative research methods generate accurate responses to sensitive questions?

When researching sensitive topics, we risk getting biased responses from our participants which can make it hard for us to generate effective evidence-based programs (see the recent learning digest on bias in research for more). Many creative methods have been proposed to reduce bias, but do they actually work? To test this, researchers tested four different survey methods, a direct question approach and three methods designed to reduce biased responses, to see which would best replicate the actual results of a ballot initiative about the sensitive political question of when life begins. By comparing the actual results of the ballot initiative to the results generated by each method, they were able to evaluate each method’s effectiveness. They found that the research methods designed to reduce bias generated responses to the survey questions that better matched the actual results of the ballot initiative than direct questions. Such methods may offer a promising approach to gathering information on sensitive issues, and can enable us to provide the resources and support that people most need.

Using creative methods to more accurately measure food insecurity

A study of food insecurity in Northern Ethiopia used one of these creative methods, the list experiment, to assess whether the direct questions that researchers commonly use may be generating biased responses. A list experiment avoids social desirability bias by enabling someone to share what they truly think or feel without worrying how other people will respond. Rather than ask people about a specific question, this method gives people a list of statements and asks them how many of the statements apply to them – not which statements, just how many. The researchers found that participants over-reported food insecurity on general questions, where participants wanted to encourage more food aid, and underreported food insecurity when the questions were about specific incidents of not having food or family members going without food, where participants wanted to protect their honor and respect. This demonstrates how direct questioning methods may not get us the information we need for effective programming, and how we can use indirect methods to enable evidence-based programming.

Do you have a study we should share for a future Facty Friday? Send an email to drg.el@usaid.gov!