If I administered a survey to participants where I had each participant rate 21 skills on their level of importance (using the same 5-point Likert scale) to a participant's job position, how should I compare the individual means of each of the 21 skills to one another (i.e., for significant differences)? Would each skill's mean be considered independent and compared using a one-way ANOVA (independent/between-analysis) or since everyone saw all the same skills, would the means be compared using a one-way repeated measures ANOVA (dependent/within-analysis)? The 21 skills are pretty dissimilar from one another and are intended to be independent constructs. Basically, should I compare means using an independent or within-participant test? Any advice would be most helpful.
I would say one way ANOVA or ANOVA coupled with some other co-variate and/or block factors (if any), than the repeated measures. What is(are) the attribute(s) of the subjects that are presumed to cause differences in the importance scores of the 21 skills? (Position? Educational background? Gender?, ....)? Whatever they are, those are the independent variables and each of the 21 skills are the responses. Which of independent variables are of primary importance to the research(ers)? - these are factors or independent variables. Those not of importance but still causing variation in the importance score of the 21 skills are co-variate (numerical, e.g. age) and should be analyzed accordingly, or if they are not required to be included in the report, but still causing variability in the importance score, they are block factors and should still be included in the analysis, but should not be reported in detail. Block factors are usually categorical (gender, educational level, position ...).
Repeated measures are usually in clinical type researches when we administer some dosed treatments repeatedly and try to see the effect over time on the same subjects. In this case, (if I understood it correctly), the survey was made just once?, no intervention at all (just observational)? No repetition on the same subjects. I do not see it to be a repeated measures. Hope this helps, sorry if there is any confusion
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA