Stores and ordered rank of products

[George P]

I have data about popularity of some products in some stores. For each of the stores I have only a ordered rank of best-selling products in them. I need help how to find information about which products are popular only in certain stores versus products that are popular in all stores.

 

[bcrider]

It depends on what your data looks like and what sort of program you want to use for analysis, but if you were to use something like the program R, you could have histograms with your y variable being popularity, your x variable being the items, and then you could partition them by stores. For an example (although slightly different because using box plots) see: http://www.r-bloggers.com/ggplot2-multiple-boxplots-with-metadata/ -with something like R, you can also limit your x variable to only show items where the popularity value is greater than some selected number for a given store.

 

[George P]

I use Excel and a little bit SPSS. However, I do not know if I should use some non-parametric test?

 

[bcrider]

I made this simple example in excel pretty quickly: http://postimg.org/image/untm61iu7/
It depends on if you need to run statistics, but if not you can pretty much do what I did assuming your columns are organized in some way. I created a file that had a column for ranking, a column for store, and a column for items. If you have something similar that needs organizing, transform those columns into a table and sort (or simply hi-light the columns and sort by itself). If you make it into a table, you can then filter your data (using 'filter') to create a bar plot only of data where the popularity is a certain value if you want; that would make things easier to look at if you have a lot of data. In the example I created there were only 7 stores and 5 items, so showing everything regardless of popularity could get pretty messy quickly if you don't filter.