why is at least one of the authors of simpleslop.es not so big on median splits? well, for one thing, it does not make much sense: why would you measure something with a certain degree of precision and then throw away much of that precision again?

... if you're doing median splits why don't you do this as well:

SiSSy is a javascript based html page that generates spss syntax for a simple slope analysis. you can indicate your variable names, their types (dichotomous or continuous), possibly the standard deviation of one of the variables and the page outputs code that you can copy and paste into spss syntax and then run on your data set.

the tool looks nice and works as intended in firefox. other browsers may mess with the layout, but it may still work, depending on the version of your browser. it does not look very nice in internet explorer. SISSy 1.1 is the most current version and it handles two dichotomous or continuous predictors and one IV.

the output looks complicated but it is not. it is commented so that you can see for yourself exactly what is done during the analyses. in this sense, sissy is also some kind of a didactic tool - once you have really worked through that code and understood what the syntax does, you should be able to write your own syntax for any other simple slopes analysis yourself.

SiSSy has been written by thomas schubert and johann jacoby and uses an idea presented first by aiken & west (1991).

SiSSy 2.0, a new version, is in the making, and will be able to handle two or three dichotomous, continuous or polytomous predictors and offer more flexibility in determining the values of each predictor for which you want simple slopes of the other variables. it will also offer complex slopes - simple slopes of two-way interactions at a specified value of the third predictor.

SiSSy can be found by clicking here