Visualizing missing data

[Missmichelle]

Hi! 

I recently distributed a survey. I am now analyzing the results. The survey has mostly categorical variables, with a few "Select all that apply", as well as some skip logic. is there a way to produce a report that will allow me to create a report of missing data? There are a lot of variables so I was trying to avoid counting the number reported as missing manually from the "proc freq". Any suggestions are welcome! I'm not sure where to start. 

Thanks!

Michelle

[ballardw]

It would likely help to provide an example of some data, not all of it but a few variables of each type and enough records to indicate the different types of "present" and "missing" data, with a way to know which is missing or not for each record.

Then rules for what you actually want reported based on that example data.

Best is to provide some example of how you intend the output to look like as well.

And you might want to provide some actual details of what "missing" means in each case. For example, if a question should not have an answer because of skip patterns that could well not be "missing". I tend to think, in survey terms, that missing is an expected response with no actual response recorded.

Also some approaches can be somewhat data dependent. There are some things much easier with numeric values than character. So are your "all that apply" questions giving a list of numeric values as codes or text strings?

Instructions here: https://communities.sas.com/t5/SAS-Communities-Library/How-to-create-a-data-step-version-of-your-dat... will show how to turn an existing SAS data set into data step code that can be pasted into a forum code box using the {i} icon or attached as text to show exactly what you have and that we can test code against.

[Ksharp · Accepted solution]

I remember @Rick_SAS has already written a blog about it .