Draft Efficient Rating Scales for Accurate Results

Posted in

By Nia Lawrence


You can afford to conduct stock surveys if you're only dealing with small populations or if you intend to distribute the questionnaires online. Customer satisfaction survey questionnaire is more complicated on larger populations, though, especially if you consider the variety of measuring mechanisms you'll use on it. Novice researchers often give in to the temptation of using stock questionnaires without even tweaking these for the intended respondents. Poorly planned surveys produce compromised results, and it isn't surprising that the implementation of measures based on these data doesn't offer any improvement or solution. Because the rating scale is the staple of a satisfaction survey, this type of research is also prone to drafting mistakes.

You can't depend on one scaling technique for all the items on your customer satisfaction survey questionnaire, since you'll be neglecting the benefits of rating scale alternatives. The Likert scale is probably the most popular technique because it's very accessible for respondents, researchers, and clients. You can even guess the response trends if you casually browse the answers. The Likert five-point scale is as straightforward as it gets, but you'll have to explain the items which use involved scaling techniques (such as the MaxDiff and Constant Sum scales). It's best to group questions into separate sections, based on the scaling techniques you'll apply on these. You'll make the survey and interpretation process easier for your respondents and clients.

Once you've decided on the type of rating scales you'll use, you have to make sure that the formats are consistent all throughout your customer satisfaction survey questionnaire. You can apply a Likert scale in a variety of ways, but its most popular format is the five-point rating scale. If you're going to apply this to one item, then you'll have to use it on all the others. Don't shift from a five-point scale to a four-point scale within the same section if you don't want to break your respondents' momentum. They'll finish the survey quickly if you sustain their train of thought, so you should also keep the labels on the points constant. If the first scale is labeled with terms like 'regularly' or 'rarely', then the succeeding items should follow through with the same terminologies.

Now that you have a rough idea of the appropriate response formats, you'll have to go back to basics and review your survey questions. If you use paired synonyms or related terms in the same phrase, then you've likely composed a double-barreled question. These questions usually have conditional or inclusive connectors like 'or' and 'and', and you'll split your respondent's answers to either alternative if you don't revise these. Your customer satisfaction survey questionnaire should only ask one question at a time, and it has to be restricted to a single concept to avoid misleading or faulted responses.

Your questionnaire will go through countless redrafting and tweaking before you come with the final draft, but it'll still have a couple of severe faults that you can only sort out through sample surveys. The rating scale is an indispensable tool for satisfaction surveys, though, so it's important that you sort out the scaling issues first. At the very least, you can perform a couple of test surveys on a sample population with your final draft, and then, polish it off with the finishing tweaks.




About the Author: