Transform a variable from discrete to continuous


New Member
Hi all, I am new on this forum so I hope this is the right section to ask questions.
Here is my question:
I have a 4 groups of subjects (between factor) repeating the same task for 10 times in different blocks (within factor).
I want to analyse differences between groups.
What i did was to perform a 4 x 10 mixed design ANOVA and look at possible main effect of "group" or interaction "group x block"
However, since I expect a linear increase of performance across block, I would like to transform the variable block from discrete to continuous.
I use StatSoft - Statistica for the analysis. Does anyone know how to do this step by step?
Thank you very much in advance


Fortran must die
I am aware of no way to make a discrete variable categorical (well there are ways such as combining likert variables into a single scale, but those are special cases. I don't know if that makes sense ). If you could convert discrete variables into interval variables everyone would do it and never do logistic regression.

You can convert categorical variables into a series of dummy variables and use those in ANOVA. Generalized Linear Models will work with dummy variables although they are not truly interval in nature.


Ambassador to the humans
I disagree with noetsi for the most part. I don't know Statistica though so can't help with the software related aspect.

Assuming that you know the order the blocks are in and aren't just assuming that there is some ordering that makes sense for a linear increase then that should be completely possible.


Fortran must die
Dason how do you convert a discrete variable to a continuous one (and if you can why would you ever run logistic regression since you could convert the data to a continuous variable and run linear regression)?