What is the social desirability effect? It’s actually pretty intuitive. We all want to appear smart, funny, attractive, and desirability. That’s just human nature. Who wants to the world to think you are an ugly loser? Certainly not me! Let’s say you ask an acquaintance if they have ever had a sexually transmitted infection. They say no when in fact they have. This is a real world example of social desirability.
How does this play into research? In order to evaluate our research results, we must consider how honest people have been in answering our questions. This is particularly relevant when we ask questions about such sensitive topics as sexuality, money, and drug use. However, we never know what types of topics our participants may be sensitive about or may “hit a nerve.” I did a study of woman age 65 years and older, and I realized that asking about finances was an issue vulnerable to social desirability in this population.
Even if a survey is completely anonymous, I would guess a group of college students who were asked to report their history of using substances such as cocaine and marijuana would underestimate their usage. With sexuality, some may overestimate their number of sexual partners (most likely men) because they are embarassed about a lack of sexual activity. Others may say they’ve had fewer partners than they actually have had.
How can we minimize the social desirability effect to maximize our chances of having accurate results? The first key is to ensure anonymity when possible. If I want to know about someone’s sex life, I will get a more accurate report if I have them fill out an anonymous questionnnaire than if I interview them face-to-face or even using the phone.
Although there are many advantages of interviewing someone face-to-face in research, it’s obvious that if someone if embarassed about something they are likely to hide that from a real person who is sitting right in front of them, no matter how non-judgmental that person may seem. People are more honest when they feel they are not identifiable.
In addition to ensuring anonymity, we can make sure participants know that other participants, as well as researchers, will not associate them with their responses. For example, we can spread out participants while they fill out questinnaires rather than having them sit close to each other. We can also provide envelopes for participants to put their questionnaires in after they are finished.
Participants may also feel more anonymous in larger groups. Even if I ensure anonymity while someone is filling out a questionnaire, they won’t feel very anonymous in a group of three! I can also make people feel less anonymous if I ask them personal questions that could identify them within a group, even if that group is relatively large. For example, if I ask 200 people in an auditorium to complete a questionnaire and only one person is African-American, asking people to tell me their ethnic category has just made one person very identifiable!
Beyond trying to minimize the social desirability effect in our research, it’s also important to acknowledge its possible presence. If I were to ask women to identify their weight on a survey, I’d have to consider that most women probably round their weight—and they don’t round it up! No matter what I do, I cannot eliminate the possibility of the social desirability effect in research.