I’m trying to create an average of percent of top three ratings on a five point scale among multiple variables in data from a survey that we conducted. Would I assign a binary rule to the scale in the particular questions? In other words would I make 1 and 2 equal 0 and 3,4,5 equal 1 and divide them to create a percent of top three box in the scale and then average those to get a composite average? Is there an easier way to do this? Thanks
Hey @btwood ,
My suggestion is:
1. Transpose tool to pivot all of your survey data into one column
2. Formula tool to create the condition to your whole dataset
3. Sort and sample to get the top three (remember that you need to have a rowID to group when sampling)
4. Cross-tab tool to pivot your data back to its regular structure
Let me know if this works. Feel free to share a sample of your input if you need an example.
Best,
Fernando Vizcaino
User | Count |
---|---|
106 | |
82 | |
70 | |
54 | |
40 |