I am analysing data extracted from questionnaires based on Likert Scale. Can anyone suggest any way of analysing the Likert Scale with the help of SPPS?
Hello Jelena. Have you seen this discussion thread started by David L Morgan?
- https://www.researchgate.net/post/How_can_one_help_people_create_simple_scales_from_Likert-scored_items
HTH.
You can perfor chi square tests to compare if two questions are related, or use ordinal regression models,non parametric tests etc.
You can perform ordinal regression in the Generalized Linear Models dialogoe of SPSS. If you have repeated measures, use Generalized Estimating Equations.
Hello Jelena. Have you seen this discussion thread started by David L Morgan?
- https://www.researchgate.net/post/How_can_one_help_people_create_simple_scales_from_Likert-scored_items
HTH.
Likert scale gives an ordinal outcome based on frequencies, therefore non parametric tests are applied, chi square can solve the problem else ordinal regression tests can be applied too on SPSS
Liket Scale data does not meet the parametric assumption regarding data characteristic. Thus you can only analyze it using non-parametric method such as Mann-Whitney, Kruskal-Wallis test, Sign-Test etc but not Anova Regression, t-test, F-test, Z-test etc.
I wonder if those who are proclaiming that parametric procedures must not be used are considering the important distinction Likert scales and Likert-type items.
- http://www.john-uebersax.com/stat/likert.htm
Only extremely conservative statisticians (in my view) would prohibit the use of parametric procedures when the DV is a true Likert scale. HTH.
I agree with Bruce Weaver, and think the Uebersax link is very instructive.
I'll also offer this paper, which I think presents a good discussion. http://ijcar.net/assets/pdf/Vol3-No2-February2016/02.pdf
It seems to me that Likert scales, composed of several Likert items, would need to be interval in nature. If the underlying Likert items were considered ordinal, then you wouldn't be able to sum their responses into a composite scale in the first place. You can't sum or average ordinal categories.
Typical Likert scale is a 5 or 7 point ordinal scale used by respondents to rate the degree to which they agree or disagree with a statement.
In an ordinal scale, responses can be rated or ranked, but the distance
between responses is not measurable. Thus, the differences between ‘‘always,’’ ‘‘often,’’ and ‘‘sometimes’’ on a frequency response Likert scale are not necessarily equal. In other words, one cannot assume that the difference between responses is equidistant even though the numbers
assigned to those responses are.
Non-parametric methods are widely used for studying populations that take on a ranked order.
In your case also, it's better to use Non-parametric methods like Spearman's rank correlation coefficient, Mann–Whitney U test, Kruskal–Wallis one-way analysis of variance etc.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts