Proper scale design can be crucial to your marketing research questionnaire. Poorly designed scales can lead to gross inaccuracies in the data your company is trying to capture.
Centrac DC is committed to properly designing scales so that we can provide your company with the information necessary to make well-informed decisions. Our experience with surveys in conjunction with leaders in the marketing research field has helped us compile information to help your surveys get better.
Here are some helpful tips when trying to design a question with scales:
Generally 5 to 7 point scales work best. If you provide too little choices, respondents may not fit that particular attribute. Too many become ambiguous.
Middle alternatives (neither good nor bad) can be dangerous to a survey. It is better to have the respondent try to identify with a particular attribute.
Numbers alone on a scale can be ambiguous. An unlabeled scale requires the interpretation of a respondent. Additionally, it can be difficult to do numbered scales over the phone.
One thing to consider in a scale is that respondents assume equal spacing in a scale is intended, therefore label your scale appropriately.
A “Don’t know” option can be included for a number of reasons including, but not limited to, unclear questions, not enough information to provide an answer, no response matches a respondents judgement, avoiding to reveal an embarrassing attribute, but according to recent estimates only 20% of people who choose the option “Don’t know” to an item today will say DK to an item tomorrow.
Label all points on rating scale when no natural metrics exist.
Never use agrees or disagrees, true or false, yes or no on a scales question.