I am conducting a research on employee turnover. therefore I have prepared a questionnaire with 5 point likert scale. there has several statements related to salary, top management support, working environment, overtime, promotion facility etc. I have also added a statement named- "i am dissatisfied with my current role". now i have measured correlation coefficients between the statement about dissatisfaction and others statements, so that i can identify which factors (i.e. statements) are more responsible for employee's dissatisfaction. now my question is how much lower value of spearman rank correlation coefficient will be acceptable to identify a statement as a reason for employee's dissatisfaction for my study?
thanks in advance.
The acceptable range of Spearman rank correlation coefficient for Likert scale data analysis is -1 to +1. This is because Spearman's correlation uses ranks instead of assumptions of normality and can be used to analyze ordinal-level data. The range of a 5-point Likert-type scale is calculated by (5 − 1 = 4) and then divided by five.
The acceptable range for the Spearman rank correlation coefficient when analyzing Likert scale data is typically considered to be between -1 and 1. A coefficient of -1 indicates a perfect negative correlation, a coefficient of 0 indicates no correlation, and a coefficient of 1 indicates a perfect positive correlation.
It is important to note that the correlation coefficient alone does not indicate the strength of a relationship, it only indicates the direction (positive or negative). The strength of the correlation can be interpreted by the magnitude of the coefficient. Typically, a coefficient greater than 0.4 or less than -0.4 is considered as a moderate to strong correlation.
There is no set range for Spearman's rho that you can consider important. A few ideas:
- Often people will use the p-value from the correlation. (Say, a p-value less than 0.05). Your software should provide this for you.
- There are guidelines for Pearson's r given by Jacob Cohen for the social sciences. These could probably used for rho also. (see the following for "small", "medium", and "large" interpretations. Note that these cutoffs aren't magical. They are just some guidelines that Cohen thought were reasonable for social sciences. https://effectsizefaq.com/2022/01/23/cohens-effect-sizes/
- It's also helpful to determine the confidence intervals for rho.
Mayur Wanjari , what does the following mean ? "The range of a 5-point Likert-type scale is calculated by (5 − 1 = 4) and then divided by five."
Do you mean the range for Spearman's rho ? If so, does this comport with the following example ?
A = (1,2,3,4,5)
B = (1,2,3,4,5)
cor(A, B, method="spearman")
Akoglu H. User's guide to correlation coefficients. Turk J Emerg Med. 2018 Aug 7;18(3):91-93. doi: 10.1016/j.tjem.2018.08.001. PMID: 30191186; PMCID: PMC6107969. @Md-Safin
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS