What is the best way to combine data from 5 columns into just one column? Of the 5 columns, only one has data in it, so there's only one response per person to aggregate into one column. I thought I could code the string variables into numbers and then add them in a separate column, but that isn't working. Any thoughts? I have mostly string data but will code them into numbers, so I can try it with either type of data. When I code them into numbers I get an unfamilar message that says "Warning: String variables being recoded differ in length. The recode may be inappropriate for some variables," but it looks like it recodes just fine. I'm recoding a Likert scale like I always do, so it seems like it came out of Survey Monkey a bit wonky. Thanks for any ideas you have! I thought this would be easy.

You may want to use the recode method to recode your Likert scales to numbers -- do you mean "5-Very high" when you refer to the survey results? If so, you can also just strip out the first character and recode it to a number.

Just check the bottom of the recode pop up screen to make sure that you are giving the new variable enough characters -- if you are recoding to 100 as your top rank, for example, make sure your new variable length is set to 3.

Thank you for your response. The variable widths all differ but the smallest is 5 and the new codes are single digits, so that should be fine. I had tried increasing the width, columns, and alignment but that didn't work. 

Hi Jon, thanks for your reply. I tried your method in syntax with several different sets of variables and one of them created a Z column but the others just gave a variety of error messages. I must be missing something. 

VARSTOCASES would also work

I'm not familiar with that, can you tell me a little more about how to use it? Thanks! 

The spss tools site has some examples.

https://www.spsstools.net/en/syntax/syntax-index/restructure-file/varstocases-and-casestovars/