Social Desirability Bias

Why this matters for your decision

Many offers from large consumer companies touch on values: climate protection, health, providing for the family, conscious consumption. This is exactly where the temptation is greatest to present oneself in a survey as one would like to be. This rarely happens on purpose. People often describe their self-image rather than their behaviour.

For a decision, this means: the more strongly an offer is linked to a socially positive attitude, the more likely a survey is to overestimate demand. A business case based on such answers plans with customers who do not exist in those numbers. Conversely, offers that people are reluctant to admit to, such as a budget package instead of the premium option, are underestimated.

The effect depends strongly on the survey mode. It is larger in face-to-face interviews than in anonymous online surveys, but it does not disappear there, because self-image works even without an interviewer present.

Example

An energy supplier asks whether customers would pay five euros more per month for a green electricity tariff. 45 percent agree. In the behavioural test, visitors to a realistic tariff page see either the standard tariff or the green tariff with a surcharge, only one variant per person.

The measured sign-up intent for the green tariff is clearly below that for the standard tariff. The surcharge is therefore viable for a smaller part of the target group than the survey suggests. A follow-up question for the next test would be at what surcharge the difference becomes smaller.

How it differs

Demand characteristics align answers with the presumed expectations of the researchers, social desirability bias with social norms. Acquiescence bias is a general response style of agreeing with statements regardless of their content. The attitude-behaviour gap describes the outcome, the gap between attitude and behaviour. Social desirability bias is one of its causes, alongside price, habit and effort.

How to deal with it

In surveys, anonymous self-completion, indirect questioning techniques and neutral wording help. They reduce the effect but do not eliminate it. In a Painted Door Test there is no question on which someone could make a good impression. People see an offer with a price in their usual online environment and choose it or not. Horizon measures purchase intent or sign-up intent, without a panel, without incentives and without anything being sold.

Limitations

Behaviour can also be socially shaped, for example when a brand promises status. In that case, however, it is part of the real market situation and not a measurement error. A behavioural test shows what is chosen, not why. Anyone who wants to understand which values lie behind a choice also needs qualitative research. And for offers that concern very sensitive areas of life, the design of the offer page and the choice of target group still need careful alignment.