More on Honest Survey Answers

In the last article, we discussed how convincing people to give honest answers may actually cut out the extreme answers more, because extreme answers are usually a result of emotion, since few people can say with all honesty that something is perfect. As we mentioned, there may not be a perfect solution to this problem, but some ideas of how you can reduce the effects of this phenomenon are discussed below. Larger Scale The larger your scale is, the less of an effect that not selecting the extreme answer should conceivably have.

A 1 to 10 scale shouldn’t be quite as affected by non-extreme answers as a 1 to 5 scale, because if people are unwilling to choose “5”, the “4” would encapsulate every single above average answer, whereas on a 1 to 10 scale, those that don’t want to choose the extreme answer can still choose “9,” while those that were never going to choose an extreme answer can choose 7, 8, or 9. Explain the Answers The “extreme” answers do not necessarily have to be “this product is perfect.” You may want to choose a scale that emphasizes that 10 out of 10 is “as good as can be right now” or something of that nature. It may be difficult to do, and may cause confusion, but it could help in decreasing the likelihood of central tendency.

Follow Up When someone chooses the answer closest to the extreme (such as 9 out of 10 or 2 out of 10), you can also ask them a follow up question that helps you find out if they are simply providing that answer because they do not want to choose the extreme answer or they’re choosing it because it truly represents their views. You can ask a question that has them choose how close to perfect the product is, and why they selected the answer. This may not be feasible for large surveys where many 9’s could make the survey last for hours, but for short surveys it may be a possibility.

None of these may solve the problem, but if you notice central tendency is occurring often with your survey, this may be a way to help solve it. You always want honesty in your survey responses, but stressing honesty may lead to less extreme answers. If you’re concerned this may be the case, the above options may help.