Right now I am wondering what I might be possibly doing wrong with my SPSS syntax. I have got 2 already dichotomized variables and I would need to combine them to make the 3rd one.
If var1 and var2 = 0, than it should be 0, and if var1 or var2 = 1 (or both of them), then it should be 1.
I tried this syntax (and many more :D), it is just not working.
if(roboSV_19_more = 0 and roboSV_19_less = 0) roboSV_19_together = 0.
if(roboSV_19_more = 1 or roboSV_19_less = 1) roboSV_19_together = 1.
I would be very greatful for any of your help, tipps and tricks. Thank you!
Hi Katerina,
Did you first establish the target variable (var3, or roboSV_19_together)? If not, you'll get error messages. And, are you sure that the two contributing variables are actually coded as 0/1?
Try something like this (and, yes, statements #2 and #3 could be combined):
compute roboSV_19_together = 0 .
if (roboSV_19_more = 1) roboSV_19_together = 1 .
if (roboSV_19_less = 1) roboSV_19_together = 1 .
execute .
Good luck with your work.
Hi David,
thank you so much for your answer! I will try it.
Have a nice day
You can use the MAX function to get what you want:
COMPUTE roboSV_19_together = MAX(roboSV_19_more, roboSV_19_less).
If both original variables = 0, MAX=0. If either or both original variables = 1, MAX = 1.
David Morse asked if you have verified that the input variables take only values of 0 or 1. If you need to make your code more bullet-proof to guard against variables having other values, or missing values, you could do something like this:
NEW FILE.
DATASET CLOSE ALL.
DATA LIST LIST / v1 v2 (2F1).
BEGIN DATA
0 0
0 1
1 0
1 1
9 9
0 9
1 9
9 0
9 1
END DATA.
MISSING VALUES v1 v2 (9).
* To make the code more bullet proof, do the computation
* only if both input scores are valid and within range.
IF ANY(v1,0,1) & ANY(v2,0,1) v3 = MAX(v1,v2).
FORMATS v3 (F1).
LIST.
Output from LIST:
v1 v2 v3
0 0 0
0 1 1
1 0 1
1 1 1
9 9 .
0 9 .
1 9 .
9 0 .
9 1 .
Good day, Katerina Pithova
In such situation I would rather:
- to use some office program (Libre OfficeCalc or MS Office Excel 365). They both have my favorite function IFS https://support.microsoft.com/en-us/office/ifs-function-36329a26-37b2-467c-972b-4a39bd951d45
- by the way, in addition, I'll consider transforming two dichotomous scales into one categorical scale with four grades.
Best regards.
Maksim Sokolovskii, I am surprised that you prefer to use a spreadsheet for a task like this. Using code in a statistical software package has the great benefit of documenting exactly what was done and making it reproducible.
Bruce Weaver Commonly, we provide massive surveys in google-forms. So at least we have:
- a file with raw answers in google-forms;
- after that we should apply some filters and partly manually check technical mistakes (wrong ID and so on);
- a file (in MS Excel or LibreOffice Calc), in which we transform raw answers into scale values;
- data file for statistical package;
- files with statistical results.
So commonly we don’t need to place some transitional data into the data file for the statistical package.
Hi Katerina
To create a new target variable out of existing variables in SPSS, use the following procedures from the Menu bar: Click Transform> Compute Variable...>
In the Compute Variable dialog box you have text boxes, options, and Function Group to aid you built any type of executable expressions.
For example, from test1 and test2 (pre/post scores) you can use above procedures to create a new paired DIFFERENCE variable ideal for identifying outliers and normal distribution of data sets.