One of the most common scaled-response format questions in survey design today is the Likert scale. It was developed by the American educator and organizational psychologist Rensis Likert in 1932 as an attempt to improve the levels of measurement in social research through the use of standardized response categories in survey questionnaires. Over the years, Likert’s original 5-point scale has taken on many new forms which we will address in the following white paper.



A commonly used 5-point Likert scale format to measure satisfaction is:

1. Very satisfied
   
   
2. Satisfied
   
   
3. Neither satisfied nor dissatisfied
   
   
4. Dissatisfied
   
   
5. Very dissatisfied
   
   

Another version of the scale which can be found in some satisfaction surveys is the 6-point Likert scale which reads:

1. Extremely satisfied
   
   
2. Very satisfied
   
   
3. Somewhat satisfied
   
   
4. Somewhat dissatisfied
   
   
5. Very dissatisfied
   
   
6. Extremely dissatisfied
   
   

In February 2006, Infosurv conducted a forum of market researchers to understand their preference between 5-point and 6-point Likert scales. Our conclusion is that most modern researchers agree that the neutral rating in a 5-point scale is needed when conducting survey research.

Of the researchers who participated in this discussion, 71% expressed a preference for 5-point Likert scales, 12% preferred the 6-point scale, and 17% were neutral on the matter. Those researchers preferring the 5-point scale cited the following reasons:

Survey respondents might truly feel neutral about a given topic, and presenting to these respondents a scale without a neutral midpoint can introduce respondent bias as respondents are forced to chose a more positive or negative response. Some researchers point out that in many cases respondents will accentuate the negative in an experience.

Neutral is a legitimate opinion that exists among respondents. Generally speaking, if we solicit every opinion of the people that are surveyed, the neutral rating needs to be included in the scale. If we are not interested in the neutral opinion, we don't have to include it in the scale.

With a 5-point scale you have a nice midpoint. The 3 rating is right in the middle and it indicates neutrality or mixed satisfaction. When calculating the mean weighted average you have a standard point of comparison. You will know instantly that an average rating of 3.4 is above neutral and a 2.8 is below.

Those researchers preferring the 6-point scale cited the following reasons:

They prefer to have an even number of ratings in the scale to have respondents commit to either the positive or negative end of the scale. These researchers disagree with giving the respondent a neutral or ambivalent answer choice.

also argue that neutral answers are rare anyways because in the majority of the cases, only those who had a positive or negative experience/opinion will want to participate in a research study.

Some researchers involved in this discussion pointed out that in the questionnaire design process, researchers must factor in respondents’ knowledge of the topic at hand. A lack of respondent knowledge may lead to an abuse of the endpoints of longer scales resulting in lower reliability than with the shorter scales. If a respondent is very familiar with the subject, for example a student rating a professor’s performance, a neutral rating may not be as necessary compared to a situation where you’re asking the student to rate his school’s financial aid policies. It could be argued that in the latter case the respondent could truly have a neutral attitude towards the subject at hand.

Conclusion

As the results of the Infosurv forum show, there are arguments for and against the various forms of the Likert scale. Though the majority of modern market researchers prefer a 5-point scale, it is ultimately the responsibility of the survey questionnaire designer expert to decide upon the scale that best fits their needs.

It is important to remember that the ultimate purpose of market research is to uncover unbiased answers that can lead to actionable results, and most research experts agree that an odd-numbered scale is better suited to capture the unbiased sentiments of survey respondents.