topic Re: PROC GLM in Statistical Procedures

PROC GLM

RaquelAR — Wed, 28 Apr 2021 23:52:23 GMT

Hi all,

I am conducing proc glm and am confused about the difference between Type I and Type III.. not sure which results I should be looking at.. any help is greatly appreciated!

Re: PROC GLM

Reeza — Thu, 29 Apr 2021 01:41:58 GMT

https://stats.stackexchange.com/questions/20452/how-to-interpret-type-i-type-ii-and-type-iii-anova-and-manova

Depends....

Re: PROC GLM

RaquelAR — Thu, 29 Apr 2021 02:09:59 GMT

Hm.. for my data I am comparing the survival of mosquitoes that are three different strains in three different levels of treatment (0 insecticide, 1 insecticide, 4 insecticides). If looking at Type III SS: both treatment and strain are significant in survival? Does the interaction variable (that is <0.0001) signify that significant of Treatment is dependent on strain?

Re: PROC GLM

PaigeMiller — Thu, 29 Apr 2021 10:42:13 GMT

This doesn't seem like a situation where Type I is appropriate. It seems like a situation where Type III is appropriate.

Does the interaction variable (that is <0.0001) signify that significant of Treatment is dependent on strain?

It means there is a statistically significant interaction. This means that the effect of treatment changes due to the level of strain, and vice versa.

Re: PROC GLM

SteveDenham — Thu, 29 Apr 2021 13:34:55 GMT

To add on to @PaigeMiller 's reply, the main effects being significant isn't really important as the interaction is significant. For a design with two factors A and B , the main effect of A tests whether the levels of A differ, averaged over all levels of B, and similarly for B, but since you have a significant interaction, averaging over the levels misses the key fact that the differences between levels of A depend on which level of B you consider and again similarly for B.

SteveDenham