Hello Sophie Lee. I am struggling to understand what your data file looks like. Can you possibly create a small dataset that shows its current structure? (You could simplify it a bit by including just one of the nominal variables and one of the scale variables.) A second file showing what you want it to look like would also be very helpful. I suggest that upload CSV files, which will be readable by everyone, even if they do not have SPSS. Also, include variable names in the first row.

Btw, what is a "letter change"?

you could work with two variables:a quantitative variable called scale, and a qualitative variable called pre and pro.

In SPSS, you can merge numeric scale variables by computing a new variable using the “Compute Variable” function and applying operations like averaging or summing without altering the original values or counts, ensuring consistent data interpretation.

In SPSS, you can merge numeric scale variables without changing their values by using the “Compute Variable” function to create a new variable that aggregates existing ones (e.g., through averaging or summing), preserving the original variable counts and distribution.

you can put the two nominal variables in one column without the actual data merging using the SPSS program. You can do this by creating two columns in the SPSS . For example, the first column is a column that is a variable with only two values A B. For example, A symbolizes the car’s speed before and B symbolizes the car’s speed after. The second column is where the two variables are combined (the car’s speed before and after). To explain more, let’s have four values that represent the speed of the cars before let’s say :60 70 100 80, and data for the speed of the same cars after let’s say: 60 90 80 75

this way make combine the pre and post scale data into one without the actual data merging

Sophie Lee , have any of the responses so far helped you solve your problem? I now find myself wondering if you are asking how to restructure your data from WIDE to LONG, as shown in this video:

• https://www.youtube.com/watch?v=hOQ4tti5MrE

Where Mike Crowson clicks on Okay to execute the VARSTOCASES command, I would recommend exiting via the PASTE option instead so that you get a copy of the command syntax.

Your goal is to combine pre and post scale variables into one variable, but retain the ability to distinguish which values are from pre and which are from post? Am i right?

As for me the best solution for this based on my experience is to stack your data (Restructure) + Add time indicator.

First is to use restructure to stack pre and post variables, you need to convert your wide format into long formatfor example one variable called speed and another called time with values pre/post. Second is rename or recode the time variable, then last is you may use the set-up analysis, you can do T-test for compare mean speed for pre vs. post, anova for using nominal groups + time or graphs for line plots by time.