I'm trying to use the Rosenberg self-esteem scale in a mediator analysis, but I am unsure whether it's possible. This scale seems to result in a score, rather than an index. Is it possible to just calculate the mean from the Rosenberg self-esteem scale and index it, or will that change the result?
Is it at all possible to work with scores in a mediator analysis. I reckon it's not since you probably need a mean (or index) to calculate with the variables, but having never done a mediator analysis, I'm a bit unsure.
Any help would be appreciated, thanks!
Hope my question is not too confusing... :)
Hello Sascha,
I would like to help you, but I confess that I can't understand your question / problem. Might you be able to present things differently?
Robert
The Rosenberg self-esteem scale results in a score, ranging between 0 and 30. It is not a mean, and to my knowledge, thus can't be indexed. Normally, when using scales based on Likert-items, you end up with a mean, which you can index and use in multivariate analyses like the mediator analysis in my case.
So, basically I'm looking for someone who has experience with the Rosenberg self-esteem scale and knows if its result can be used in a mediator analysis, and if yes, how.
Sascha Thürmann. Thanks for getting back. Typically with scales such as the Rosenberg scale, researchers either just add up the individual responses to get a total or they take the mean of the responses. (Of course, care needs to be taken if there are any missing data.)
In essence, therefore, I wouldn't worry whether you used the total score or the mean of the scores. I'm sure you'd get the same results in your analyses either way. Put it to the test!
Maybe you should do whatever most researchers in your field do in terms of simply adding the responses or taking their mean.
- Badges
- Science topic
- Similar topics
- Indexes