I have three Likert based surveys. One measures servant leadership, one measures transactional leadership and the other measures employee satisfaction. I have received conflicting advice on how to analyze these. Servant leadership and transactional leadership are the independent variables. Employee satisfaction is the dependent variable. All three surveys were given in a single survey. It seems to make more sense to me to compare the independent variable scores separately to the dependent variable because I just want to know if there are relationships between aspects of these leadership styles and job satisfaction scores. However, I was also given advice to compare all three at once. What is the best method here or is it a matter of preference? These leadership styles are not really complementary and would seem to have an opposite effect and I feel that is why I am not confident about using a Spearman's on all three at once.
What exactly do you want to know? Are you proposing a multiple regression? What are the correlations between the two leadership surveys? Singularity might be an issue. What is your sample size?
I'm highly skeptical; what theoretical understanding would there be if both correlated with employee satisfaction, especially since both leadership surveys look at the same factor (and probably are at odds)?
In this case, I have a sample size of 100 and I’m looking to see if there
is a relationship between these leadership styles and employee satisfaction by using Spearman’s correlation because the data is ordinal. Should i just compare scores for each leadership style to employee satisfaction separately?
You can look together in correlation; just as interesting is the correlation between the two leadership surveys. Why? As theoretical constructs, presumably servant and transactional are at odds and should be in conflict. You could possibly conduct other statistical tests, but you should have a hypothesis and realize conducting multiple tests will surely report a significance somewhere which might be error.
Yes, you can conduct the statistical tests. With full demographics, the results could be informative. Good luck.
You are right, I am expecting these leadership styles to be on opposite poles, so to speak. And if they’re not, that’s going to create some problems in my study but I’m going go ahead and pursue it. Thank you for your help!
- Badges
- Science topic
- Similar topics
- Philosophy
- Moral Philosophy
- Free Will