[Laura C]

Thank you! Very, very much. All the help is greatly appreciated.

I will try it with the new names. I actually do have 20 variables, but from 10 on, I would need to extract 3 characters for the renaming. Writing 2 loops was my first thought this time.

Again, thank you.

[Laura C]

I am on a roll!

Now that I have solved the problem above, I am facing one that likely does not have a solution!

I have q1_1 q1_2 q1_3. Imagine that each of these variables has months as answers. But, say February could show up in either q1_1, q1_2, or q1_3. To explain, people were allowed to pick multiple months as answers to the initial question.

Is there a way I could get the number of occurrences for each answer across the 3 variables? For example, if January was mentioned 3 times in q1_1 and 1 time in q1_2 and 5 times in q1_3, could I somehow have a tabulation across all these 3 variables that would tell me that January was mentioned 9 times? I mean without having to manually add up, or to create countless yes/no variables, one for each possible answer.

Some other questions have longer answers, not just months.

Thank you again! So much.

[Rich Goldstein]

you might want to look at the user-written -mrtab- (use -search- to find it); you might need to re-format your data before using this command

[Laura C]

THANK YOU! This seems to work well so far. It is so great to have this! Amazing!

[Laura C]

THANK YOU! This seems to work well so far. It is so great to have this! Amazing!

[Nick Cox]

tabm from tab_chi on SSC also addresses this problem.

[Laura C]

Thank you! I got tabm now too. Both great time savers. I wish I knew them a long time ago.
In this case, for me, mrtab had a bit of an advantage because it gives the number of valid respondents and calculates the % as well.

[Nick Cox]

Yes; mrtab is more SPSS-like. tabm is just a wrapper for tabulate with no extra bells or whistles.