Likert scale has a unique way of analysis. How can I decide whether I have to use standard deviation, Pearson r, ANOVA or only frequency? How chi-square test can be performed on these data?
Hello Zuhayer. John Ubersax has a nice note on this topic you may find helpful--see the link below.
http://john-uebersax.com/stat/likert.htm
Hello,
There are so many issues in your question that I will answer only to two of them.
1. Chi square test pertain to categories that are not to be set in order (e.g. eyes' colour) and has little if anything to do with likert.
2. Up to 5 answer categories with likert format one should treat variables as categorical, from 5 on they may be treated as continuous.
Regards, Witold
Hello,
Thank you for your answer.
Regarding your first answer, if I want to see association between gender of the respondents and job satisfaction, will it be reasonable?
Absolutely,
And if job satisfaction is a continuous variable then independent sample t-test poses a viable way to see if differences across gender are statistically significant.
Witold
Hello . If you have categorical variable, it is better to use non parametric approach. if you want to find the association between them, you should use chi square test or if you want to significance the differences among groups you can use ANOVA. Other wise independent and dependent sample t-test. on the other way you can use factor analysis to validate the data set and measure the variability of the variables basically.
Hello Zuhayer. John Ubersax has a nice note on this topic you may find helpful--see the link below.
http://john-uebersax.com/stat/likert.htm
My colleagues have , in my humble opinion, eloquently and concisely answered your questions and the link from Bruce is very informative!
If you have any clarification questions that require further input, please feel free to post (I am following question and responses now and will be notified) or contact me privately [email protected]
Hello Zuhayer
I agreed as other colleagues has stated precisely that, if you want to see association between gender (categorical scale) and job satisfaction (likert scale of 3 to 5) (such as influence of gender on job satisfaction), it's best of you to use chi squared test .
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA