is it logical to categorize the perceptions using Low , Medium and High or shall we apply any other test to analyse the perceptions.
There are a couple of things you'll want to consider
1. is their a common way that this variable is categorized in research?
If so it's good to use that. It will make it easier to compare your results to that of others.
2. do your theories support a certain categorization?
3. what kind of analyses are you going to use?
A categorization of low-medium-high is not going to be a scale, so this will limit your options and complicate your findings. If your theories allow it, a dichotomous categorization (low-high) might be better.
4. what is the distribution like?
Ideally you'll want somewhat comparable group sizes, so if your distribution is skewed you can change your categorization to reflect this.
In agreedment with the suggestions of the previous answer, I would like to highlight that it is utterly important to first consider how you are going to analyze your data. All the rankings in a Likert scales represent a number code (ordinal data) of a qualitative (categorical) variable which shows the properties of the data. If you have a five point Likert scale, for example, you already know that 1 is the lowest score and five is the highest score. However, you can use your 5 Likert rankings to analyze the ordinal data without limiting the options using a low-medium-high categorization.
Maybe I can help you further if you provide more information about your research questions and the analysis you want to perform.
Meanwhile, this paper could be helpful:
Liddell, T. M., & Kruschke, J. K. (2018). Analyzing ordinal data with metric models: What could possibly go wrong? Journal of Experimental Social Psychology, 79, 328-348.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size