There are several types of response biases that can influence your results and decrease your survey validity. Using insights from psychology and decision theory, you can avoid these common types of survey bias from affecting your survey results.
In previous articles, we’ve talked about how the order of your questions can introduce bias into your results. This is also true for the order of your response options. For closed-ended questions, the first and last responses tend to be picked at a higher rate than middle options. This is due to primacy and recency effects. Primacy and recency biases are known as the tendency to prefer the first or last option provided, respectively.
To reduce response order effects in your results, ensure that your answers appear in a random order for each respondent. This is a quick fix that can easily increase the quality of your data!
The way you word a question can have a significant influence on the way respondents answer. A leading question is one where the question guides respondents to answer in a certain way. Despite their subtlety, leading questions can have significant impacts on your results.
One study  showed the effects of leading questions when it asked students to estimate the speed of a vehicle in a collision. Their findings were a hit in the scientific community (pun intended). Students’ estimates were impacted by the verb used in the question. Asking “how fast were the cars going when they smashed each other?” yielded higher estimates than when ‘smashed’ was replaced with less intense verbs like ‘hit’ or ‘contacted’.
The most effective way to avoid leading questions is to ensure that your wording is as neutral as possible.
Many people find it easier to say yes than to disagree – and this goes for survey responses too. This is called the acquiescence bias. It’s defined as the tendency for people to agree to a statement or opinion (even if they don’t feel that way). This bias is prevalent in yes/no and agree/disagree questions.
Much like a leading question, the best way to avoid acquiescence bias is to keep your questions neutral.
Having a strong understanding of the different types of biases that may affect your market research is important. Make sure that you create your questions thoughtfully and take the time to review them before you start your research.
1. Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of auto-mobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning and Verbal Behavior, 13, 585-589.