Research Tips

When using scales for responses in questionnaires, it’s important to ensure that the labels used all represent varying levels of the same thing and that the labels are listed in descending or ascending order. For example, a major organization recently fielded a questionnaire that asked how important a variety of attributes were to the respondent. The labels associated with the different choices within the scale (in the order given) were:

  1. Very important
  2. Important
  3. Does not influence my decision
  4. Somewhat unimportant
  5. Totally unimportant

The first problem with these labels is that attributes that are important and attributes that influence decisions may not necessarily be identical and yet this scale mixes the two. An attribute may be important to a particular consumer but may be equivalent for all options that he or she is considering and therefore not influence the consumer’s final decision at all.

Assuming that in this situation, importance and influence are identical, the second problem with the labels used in this scale is the order of the third and fourth items. An attribute that is “somewhat unimportant” would still have had some (albeit limited) influence on a decision, and therefore should be listed before “does not influence my decision”. And, again assuming that importance and influence are identical in this situation, it’s not clear how “does not influence my decision” is different from “totally unimportant”.

Further tips concerning the use of scales in questionnaires:


You can have new Research Tips e-mailed to you when they are published …