Likert Scale


New Member
Hi! Can you help me with my stat research problem? what is the difference between 3-point likert and 5-point likert scale? which is better to use and what is the purpose of using 3-point?5-point?

thanks..! hope you can help me. :wave:


Hi Mae,

The scales you mentioned in your post are mainly used for perceptions analysis.

A 3-point scale allows survey or interview respondents to give their opinion/perception about something based on three possible responses; namely good, average, or bad. 5-point scales (commonly known as Likert scales) give respondents more options to choose from, because they are based on five possible items to measure perceptions; that is: very good, good, average, bad, or very bad. 10 point scales operate with the same principle.

For example, in a student survey, participants are requested to give their opinion about their research methods course.

3-point scale

What is your opinion about the literature used during your research methods course:

( ) Very positive
( ) Neutral
( ) Very negative

5-point scale

The question states:

What is your opinion about the literature used during your research methods course:

( ) Very positive
( ) Positive
( ) Neutral
( ) Negative
( ) Very negative

10-point scale

How would you rate the quality of the literature used in your research methods course?

(1 very bad-10 excellent)


Other examples:

Guttman scale: Useful to assess the extent of agreement or disagreement to a given argument.

Our current education policy favors the inclusion of marginalized groups.

( ) Totally agree
( ) Agree in general
( ) Neither agree nor disagree
( ) Disagree in general
( ) Totally disagree

Scale types include Lickert, Thurstone, and Guttman scales amongs others.

From a very general standpoint, it is better to have more points in a scale because it provides you with more insight about the perception that you are trying to measure. An important consideration is that if you have a small sample of respondents in your study, using a 3-point scale will affect the validity of your findings due to the fact that 3 point samples polarize your results into items that denote very good, average, or very bad; but nothing in between.

The webiste: has lots of information about these topics.

All the best,



Fortran must die
It depends on what scale makes sense. While a 5 point scale allows more options for respondents that is only useful if respondents actually can distinguish the phenomenon meaningfully on five points as compared to three. For example if the first scale is 1 strongly agree 2 neutral and 3 strongly disagree and the 2nd 1 strongly agree 2 somewhat disagree 3 neutral 4 somewhat disagree and 5 strongly disagree, if respondents don't really understand the difference between say 1 and 2 (or can not measure it) then creating a five point scale forces them to make judgements they can't. The results are worse not better since you end up with error in your model by forcing people to make judgements they would not or could not do.

Statistically, it is better to have more levels because as you beyond 5 (meaningful) ordinal levels the variable becomes more interval like and thus you can better use many methods such as regression and ANOVA (if the variable will be dependent). But you have to trade this off against the danger noted above.


New Member
Hello all,
I am trying to analyse Likert scale data for my research too. I am not much of an expert on statistics and am really struggling with this part in my research. Can you please help me with my analysis? I have created a very simple questionnaire with 5 questions. The answers to each of those questions are in the form of a 4-point likert scale. Can you advice the best way to analyse this data? I have read so many things on the analysis of likert scale data that I am thoroughly confused now. To give you an example, I have:

Q.1 The essence of financial regulation has no strategic impact on an organization
1. Completely agree; 2. Slightly agree; 3. Slightly disagree; 4. Completely disagree.

I have a very small sample size (about 20) as I am trying to get the opinions of middle and upper level bankers on financial regulatory reforms. On getting the answers, I want to use the question statement into a null hypothesis and accept or reject the hypothesis based on my findings. I have downloaded minitab too. I'd be extremely grateful if you could you please tell me what would be the best way to do this?

Thanks a ton.