I got data on 03 likert scale ( No , Present till 50% , Present more than 50% ) of 10 core indicators. Each core inductor have more 9 sub-indicator. I scored (0/1/2) the each sub-indicator and took the means of each core indicator. Is the Means of Core Inductors Continuous Data or Nominal Data? ANOVA can apply?
Whenever we have an average of variables irrespective of nature variables for which average is required, averages like arithmetic and geometric means, etc are always continuous. ANOVA may be applied provided that it satisfies residual analysis as well as other assumptions like related to heterogeneity, etc.,
Its ok @Srijan Lal Shrestha.Is sample size play a role to consider Means as continuous data? i.e more than 300 sample size the statistical means consider as continuous data. And please mention some reference that Statistical Means are Continuous Data so that I can put that reference in my thesis. Regards
I believe what you want is ordinal logistic regression with an ANOVA model RHS. This is because your DV is neither continuous nor nominal. See the attached notes. Best wishes, David Booth PS also see the link ANOVA by regression:
https://www.google.com/search?q=ANOVA+BY+REGRESSION&ie=utf-8&oe=utf-8&client=firefox-b-1
It's not clear what data you have and how you are handling it.
If you are summing up or averaging responses from several questions, and these sums have a considerable variation, and you assume that these sums come from a normal population, then you may be able to use anova. The assumptions of anova will still need to be met.
If you are analyzing individual questions, ordinal regression is a good approach, as David Eugene Booth suggests.
The means are necessarily numeric data. They are not truly continuous, but this may not really matter. They are interval. This is because if you are taking the means, you have already made an assumption that the original categories are interval. It's not clear that it's valid to make this assumption, but you have made it when you find the means. Note that you have made a choice to assign the values (0, 1, 2) to the categories. You could have chosen different assignments, and this would affect your results. The numeric translation you have chosen may not be ideal for your categories, as your categories could be thought of as (0%, 25%, 75%), and so the assignment of (0, 1, 3) would make more sense. This is the kind of assumptions you have made when you translated your ordinal categories to numeric values, whether you realized it or not.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA