- All variables are continious (measure: scale)
- I have added 4 variables (dependent, independent, mediator and control variable)
- I'm using IBM SPSS Statistics 23
- I have used the following options: correlations (bivariate), Pearson, two-tailed, flag significant correlations, exclude cases pairwise (which are all standard options)
Hello doctor
If you are meaning how to interpret the results , first step is to do normality test to know if you will use Pearson (normal distribution) or spearman (abnormal distribution).
Interpretation of correlation
Severity : weak , moderate , strong, perfect
Positive ot negative
P value 0.01 highly statistical significance
( r value = ..... and p value = ....)
Hello Nicoletta,
I don't have access to v 23 (do use v 27), but I would presume that much of the internal workings are consistent across versions.
This syntax generates p-values and, by default, flags significant correlations at the .05 level:
CORRELATIONS
/VARIABLES=x y
/PRINT=TWOTAIL NOSIG FULL
/STATISTICS DESCRIPTIVES
/MISSING=PAIRWISE.
If your version behaves differently, I'd guess that there is a global setting that can be adjusted (perhaps in user preferences). Another possibility is that the European versions were programmed to default to .01 instead of .05. Of course, the program does furnish p-values (and you can adjust the number of decimals by right-clicking on the output table, then selecting cell properties for the p-value cell), so you can refer these values to any alpha threshold you like.
Good luck with your work.
- Badges
- Science topic
- Similar topics
- Mathematics
- Algebra
- Matrix