Which Statistics To Use?

[lstanley]

I am looking at the effects of 4 different driving training methods on 4 different groups of people. My dependent variables are thier number of accidents, number of points on driving record, and number of citations at 3 months after receiving their license and then a year later after receiving their license. What test do I use to determine if any of the training methods were effective?

Also each group will be rating their perceptions of effectiveness using a likert scale. What test would I use to determine if there are any differences amongst the four groups in the perceptions of effectiveness?

thank you much!!

 

[JohnM]

On each dependent variable, do a 1-way ANOVA.

I would not recommend MANOVA since it requires you to combine the DV's into one linear combination, and in this case it doesn't make sense to do so.

 

[lstanley]

Thank you very much for your help. I guess I am not really understanding the differences between a MANOVA and an ANOVA. Why is it not advantageous to include all the DV's into on combination? Don't I inflate my Type 1 error rate when I perform multiple one-way ANOVA's? Much thanks!! This site is wonderful!

 

[JohnM]

Technically, yes, your Type I error rate will go up, but it's not a big deal because I'm assuming that you planned, as part of your study, prior to collecting data, to examine these variables....

MANOVA will tell you which IVs are significant for a specific linear combination of DVs, which may or may not make sense from a practical viewpoint.

If you were able to combine the DVs into a meaningful, composite "index" of driver safety, then OK, maybe MANOVA makes sense.