Hi all,
I ran a repeated-measures experiment with two treatments and two kinds of participants. So I have both within and between group variables, and each variable has two levels. If I am looking for any interactions or main effects of these IVs on a continuous DV, what's the most appropriate analysis? I was thinking mixed factorial ANOVA, but wasn't sure if it required variables with more than 2 levels. And if mixed factorial ANOVA is the way to go, can someone let me know what the assumptions are? I'm getting well confused looking at all the different options and can't seem to find a source of instruction that deals directly with what I'm doing.
Much thanks for your time.
if you like to try R, i would propose the lme4 package; you can do interactions between any combination of discrete and continuous variables
http://cran.r-project.org/web/packages/lme4/index.html
Thanks all, I'm using SPSS. I'll look into those other packages but the reality of my timeframe means I'll probably stick with what I know. Cheers
Long term, switching to R is a good idea. There have been many advances and improvements that can make a substantial difference in our understanding of data (e.g.Wilcox, 2012, Intro to Robust Estimation and Hypothesis Testing) that are impossible to apply using SPSS. When it comes to taking advantage of modern methods, SPSS is hopeless. Even in our undergrad stat courses, we have switched to R. Another incentive is cost. R is free and IBM is intent on pricing SPSS out of the market.
Short term, in spss you may use: Analyze / General Linear Model / Repeated Measures... In the first dialog box, name and specify # levels of within-subject factor, click 'Define' and add the variables matching those levels (assumes wide form data entry...all levels of a repeated measures factor are entered as separate variables), then add the between subjects factor in same dialog box.
Long term, I concur that R has numerous advantages, though you often have to be cautious about methods used in specific procedures (e.g., lm uses type I SS as default computation method: see this site for additional information-- http://www.uni-kiel.de/psychologie/dwoll/r/ssTypes.php ).
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA