One thing which is popping up in my mind. In a data table, I have two groups but the difference of N numbers between the first group and the second group is not compatible. For example, can I use the Chi-square test (based on the categorical data)? or is there any way for normalization? I have attached an example of such a data table.
Hi Sourav Mazumder. In the table you attached, the cell counts you show do not sum to the group sizes. E.g., 150+70+20 > 160. I assume, therefore, that each row in your table represents a 2x2 table like the ones I have attached. Is that right? Thanks for clarifying.
Thank you Bruce Weaver sir, this is exactly what my data table comprises. One thing I should mention, at the time of data collection through web survey tool (e.g., Google forms), usually, generates charts automatically, but I have to split and code the options (in a cell) to extract the frequency of the responses of a particular option. After doing all the kinds of stuff, then I could be able to get the table that I attached earlier.
I also have attached an example (data table) that shows the raw data collection.
- Badges
- Science method