Stare at any market research survey long enough, and you'll be able to poke holes. All market research methodologies incur some type of bias whether it's response bias, sample bias, or sequence bias. It is nearly impossible to create a controlled environment which produces fully authentic results. It is the job of your market research consultant to choose the best methodology to eliminate as much bias as possible, and construct a research instrument which mitigates potential response bias.
What is sequence bias in market research?
Sequence bias is defined as prejudice or favor shown to a thing, person, or group in market research due to it's particular order in a list. For example, if you were to ask respondents what their favorite food was and pizza was always listed first in sequence ahead of burgers and hot dogs, respondents may be more apt to choose pizza because it was the first one they read. The same would be true for the last one shown (hot dogs) which may cause that food to be least likely picked because it was last on the list.
How do you fix sequence bias in surveys?
You must find a way to change the order of your responses in each survey conducted. For example, if you were to complete 100 surveys on favorite foods and each of those three items were randomized on each survey, sequence bias would be eliminated and results would be more reliable and unbiased. The availability of CATI software, online survey software, and mobile-friendly platforms has given programmers plenty of easy to implement options for order sequences in their market research study.
What options exist for reordering answer categories?
Here are 4 options to eliminate sequence bias in your answer categories on your next online survey.
- Inverse Order - alternate extreme start points on each new survey (Example: Choices A, B, C, D to Choices D, C, B, A)
- Randomization - randomize survey choices for each new survey (Example: C, A, D, B to D, C, A, B to B, D, A, C)
- Forward Order - The starting point of each new survey advances forward one answer choice. (Example: A, B, C, D to B, C, D, A to C, D, A, B)
- Backward Order - The starting point of each new survey moves backward one. (Example: A, B, C, D to D, A, B, C to C, D, A, B)
A quick reminder to keep an eye on your anchor or last choice in your multiple response questions. If you list an other(s), a none of the above, or not applicable answer choice at the end of your scale, make sure you are not including this anchor in your inverse, randomization, or reordering. These anchor categories should always be listed as the last choice for respondents.
Drive Research is an online survey company in Syracuse, NY. For more information about our market research services contact us at [email protected] or by calling 315-303-2040.