PG851: GLM--ANOVA

This blog is for the use of students enrolled in either section of PG851. It is intended to be a forum for asking, discussing, clarifying, and helping.

Sunday, February 15, 2009

Q regarding Repeated Measures Assingment

I wonder how to interpret the significant effect of between-subject factor, (which is "group" in the second analysis in our repeated measure assignmet) in a mixed design model. In addition, what further tests do I need to do with this significant result?

posted by Sun @ 1:01 AM  

3 Comments:

  • At 11:13 AM, Blogger ann said…

    interpret the significant effect of group as you would for a oneway ANOVA, which simply means that there is a significant difference between the groups with and without intervention.

    whether you need to run additional tests depends on whether the answer you obtained with ANOVA specifically pinpoints where the significant difference takes place. do you think you can clearly identify effects re: time and intervention with what you have so far? if so, then no more additional test is needed, but if not, run some post hocs.

     

  • At 8:34 AM, Blogger Sun said…

    Thank you,Ann. Then let me add one more question. I am not still clear with what it exactly means that groups are significantly different in this model. Does it mean that these two groups are significantly different accross two Time points, or it means they are different within each Time point? If they are different across two time points, how do we compare the difference with the given data set?

     

  • At 6:38 AM, Blogger Mari said…

    A significant interaction means that the effect of one variable depends on the level of the other variable.

    A significant main effect means that, averaging across all levels of the other variable, there is a significant difference between levels of the one variable.

    Because it's very possible to get lost in the "one variable" and the "other variable" language above, let's use an example.

    Imagine that you were looking at clinical skills for Fuller students and Rosemead students at the beginning of graduate school and at the end of graduate school. A significant time by school interaction would indicate that the effect of school depended on the time at which you were looking (so, for instance, that Fuller students and Rosemead students weren't different at one time point and were at another, or that one group got better over time and the other one stayed the same or got worse, or any other nonparallel line situation).

    A significant main effect of time would indicate that, averaging over both schools, students were different at time 1 versus time 2.

     

Post a Comment

<< Home