How to Develop Rating Scales

Rating scales measure many activities, from consumer satisfaction to personnel performance, academic outcomes and medical progress. Creating a rating scale provides you with a quick assessment tool to apply to your business. Using rating scales provides information on how your enterprise is functioning, failing or succeeding. Critical to creating or using a rating scale is determining exactly what you wish to measure.

Instructions

1. Define what you want to measure. Usually this is expressed in gradations of behavior: always, often, sometimes, seldom, never. You can use such a scale to collect hard behavioral data from observation: the rate at which nursing students wash their hands when moving between patients, the frequency of open-ended questions asked by teachers, or the departures of customer service reps from company-created service scripts. You can adapt scales to quantify behaviors: one to five times per hour, frequency of customer reorders, self-reports of pain experienced on medication, or customer satisfaction with delivery, packaging, complaint-resolution or product quality.

2. Define the population you wish to assess. This can range from two-income families to owners of multiple dogs to minority teens under age 15. The closer the relation of your population to the questions you ask, the clearer the information you'll produce.

3. Test your rating scale on a sample population before widespread use. Testing may reveal that questions are hard to answer, lead responders off on tangents or don't yield the information you need.

   Knowing that customers report frequent difficulty with using a product doesn't tell you anything specific about the difficulties: you may need an easier-to-remove cap design, a better applicator or better directions for use. The number of questions in your rating scale may be so high that either administrators or responders get bored or confused: "I already told you that."; "If you reorder more than 3 times a year, how many times do you reorder?" Some scales need an N/A ("does not apply to me") category; women may buy your men's cologne but can't answer a question about how frequently their partners use it. Thirteen-year-olds love playing your video game but are probably not the family members who can afford to buy it.

4. Use your sample data to tighten the focus of your scale. If you've already selected your survey population from product users, a "Never" response is irrelevant. Questions that produce unpredictable responses in your sample population may need clarification or removal. Results may not yield information that's truly useful. Use repeat sampling to focus your rating scale on the information you need.

5. Establish population criteria that are large enough to get the information you want. Recognize that many people find surveys or rating scales something they don't want to participate in. Busy shoppers may be too rushed to answer lots of questions; a product sample may slow them down long enough to get your information. Much as you want to know the income levels of your customers, they may regard tightly focused questions as excessively intrusive. Senior citizens who've been warned repeatedly about dishonest phone solicitors may hang up without listening to what you want. Make your population large enough to allow for some refusal to participate.

References

• University of Southern Florida: Classroom Assessment
• Creative Research Systems: Survey Design
• Alamo Community College District: General Guidelines for Creating Surveys