Beware of These 3 Common Biases
By Julia Ford

As researchers, we know biases are nearly impossible to avoid. But recognizing them and responding accordingly can make all the difference in getting the most accurate and actionable quantitative results.

Here are 3 common biases we encounter regularly that you should consider before your next market research survey.

Question Order

When taking a survey, respondents will often carry references from question to question. For this reason questions will often influence responses to the questions that come afterwards. This can be positive in cases where upfront questioning is used to get the respondent warmed up in the category you¹re asking about prior to diving deep. However, in other cases certain lines of questioning could cause unwanted influence on what you¹re trying to capture (e.g., asking unaided awareness after exposing the respondent to a question containing a list of brands). Make sure your questionnaire order is logical and works for what you’re trying to accomplish.

Beware of these 3 Common Biases

Inability to Answer

When designing a questionnaire it is always important to ask yourself would I be able to answer this? This practice helps avoid writing questions that are so difficult that respondents often guess their response or give a grossly inaccurate answer (e.g., How much did you spend on beverages in 2014?). Ensure your findings are based on real data and avoid bogus answers by writing questions that you can reasonably answer yourself.

Self-Selection

This occurs when survey respondents are given the option to respond to a survey or not, which is the case in virtually every quantitative survey we field. We notice the effects of this most when using client provided sample. It has been our experience that responders of client sourced survey tend to be the more emphatic or heaviest users, leaving us with a disproportionally low amount of less frequent or lapsed users. We often correct this by weighting the data to match the actual proportion of user types within the client’s database.

What other biases do you encounter and what do you do to avoid them?  Comment below.

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.