Version 1.12.5 fixes the use of the shortcut "COMP" - In SPSS version 21 (at the latest, it may be the case in earlier versions, too), "COMP" is no longer unambiguous, it could mean both "COMPUTE" as well as "COMPARE DATASETS". Therefore the shortcut has been expanded to "COMPU". Granted it almost isn't worth abbreviating anymore, but I will keep the short version. If you run into any troubles because of this, try the long version.Version 1.12.4 fixes two errors in the variable naming syntax portion (thanks to katharina s.!) These errors were inconsequential for the estimation procedure itself as well as the interpretation.Version 1.12.3 fixes an error in the last portion of the syntax introduced by
"to check results with http://www.people.ku.edu/~preacher/interact/mlr2.htm.". The error was a missing "/ORIG"
switch in the regression model estimation command (Thanks to Latisha who caught the error!)
This error was inconsequential for the syntax before this section, this section adds no information that is not contained
in the sections before that. this section only serves to cross check results with a different but equivalent calculation
wherein the error variances of the simple slopes tests are calculated by hand. Also, the link to http://www.people.ku.edu/... is defunct, use [http://quantpsy.org/interact/interactions.htm] instead.
Version 1.12.2: Please note that version SiSSy1.12 contained a serious error. I am very, very sorry for that.
The simple slope that the syntax generated for MODERATOR = -1 (for dichotomous moderators) or MODERATOR = -1SD (for continuous variables)
was the same as the slope for MODERATOR = +1/+1SD.
Also, the syntax was missing periods at the end of two commands.
Both of these errors are corrected (hat tip to Sandra and Jan who noticed these mistakes first!). So, if you have run analyses
using SiSSy1.12 (between August 11 and August 30, 2011) please rerun them using SiSSy 1.12.2.