I have asked SPSS for the descriptive statistics using analyse, descriptives, explore as I need the median, and interquartile range. I have over 200 variables that I want to get this for individually so I can create a new data set.
In the explore box I move across 1 variable at a time to the dependants list, (I do not add any other information to the other boxes,) I get for example:
Descriptives
Mean 704.31
Median 670.50
Interquartile Range 252
However, as this would take forever, I add the variables approx 50 at a time, ( and can still get the results for the individual variables. I get the following as the result of the same variable.
Mean 699.75
Median 715
Interquartile range 237
This should give me the same results for this one variable surely, or is something going on that I am unaware of?
Thank you
Dear Emma,
That sounds weird. The only thing I can think is that somehow putting several variables into the dependent list affects the number of valid cases for each variable.
Hi Emma
there should be no difference, the explanation of @Peter is very possible. The exploration option is based on the number of valid cases and analyzes for all the variables based on the one with the least value loss.
the descriptive option, considers analysis in function of the valid cases by variable, for that reason, it is important to show "n" for each variable.
Note the n of both tables, it should be different.
Peter & Carlos are right: The default setting for the MISSING sub-command is LISTWISE. Try changing it to PAIRWISE.
https://www.ibm.com/support/knowledgecenter/en/SSLVMB_25.0.0/statistics_reference_project_ddita/spss/base/syn_examine.html
Thank you so much Peter, Carlos and Bruce. This has resolved the problem, now to go back and re run my analysis.
Emma
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS