Hello,
I would appreciate your helping recoding multiple variables into a new variables with 4 levels but using certain conditions for each level. I have 2 variables, (VigPA, ModPA), each with a range of 0-1260 mins, and I need to recode them into one new variable (PAgroup) with specific conditions for each level.
That is:
Level 4 (Highly Active) will include participants which performed more than 300 mins of ModPA or more than 150 mins of VigPA.
Level 3 (Active) will included participants which performed between 150 and 300 mins of ModPA or between 75 and 150 mins of VigPA.
Level 2 (Insufficiently Active) include participants which performed between 0 and 150 mins of ModPA or between 0 and 75 mins of VigPA.
Level 1 (Inactive) include participants which performed 0 mins of ModPA or 0 mins of VigPA.
I've tried writing my own syntax and Im able to successfully code the conditions for level 4 but not for the rest. I've included my code below for reference. This is my first time using SPSS and would appreciate any advice on how I can effectively achieve this goal. Thank you in advance for your time!
DATASET ACTIVATE DataSet1.
DO IF (modPA >= 300 | vigPA >= 150).
COMPUTE PA_TEST=4.
ELSE IF (150
Dear Alejandro Gallego ,
That looks painfully difficult! I'd recommend downloading into Excel and doing the calculations there step by step then loading back into SPSS. Once you have got it working you can try to establish an SPSS compute equivalent, maybe building it up step by step.
Alejandro,
Instead of : ELSE IF (150 = 75 & vigPA < 150)).
Also, you might try initializing COMPUTE_PA_TEST = 0. to handle cases where you are missing a value. Without that initialization, SPSS apparently sees a missing value, scratches its head over the DO-IF, gives up and goes to the next case, skipping all your fine-tuned IF-ELSE logic, per IBM, as I understand it. See this link https://www-01.ibm.com/support/docview.wss?uid=swg21476060
Peace,
Fran
Hello Alejandro,
Note that your Level 1 and Level 2 definitions both include 0 minutes. Levels 2 and 3 duplicate the 150 / 75 thresholds.
Here's a simple way to proceed, which does include a nil category for missing cases (PA_TEST = 0), and corrects the duplicate values.
COMPUTE PA_TEST = 0 .
IF ((modPA >= 0) OR (vigPA >= 0)) PA_TEST = PA_TEST + 1.
IF ((modPA > 0) OR (vigPA > 0)) PA_TEST = PA_TEST + 1 .
IF ((modPA > 150) OR (vigPA > 75)) PA_TEST = PA_TEST + 1 .
IF ((modPA > 300) OR (vigPA > 150)) PA_TEST = PA_TEST + 1 .
EXECUTE .
MISSING VALUES PA_TEST (0) .
EXECUTE .
Good luck with your work!
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS