GLM or one way ANOVA for withing groups data?

Posted by pegah on Aug 02, 2013; 9:45am
URL: http://spssx-discussion.165.s1.nabble.com/GLM-or-one-way-ANOVA-for-withing-groups-data-tp5721455.html

Hi,

i have problem with data analysis method.

my DOE:

 3 factors with 2 levels for 2 of them and 4 levels for the other one (i.e. 16 treatments in 4 groups)
group one:
1 1 1
1 1 2
1 1 3
1 1 4
group 2:
1 2 1
1 2 2
.
.
group 3:
2 1 1
2 1 2
.
.
group 4:
2 2 1
2 2 2
2 2 3
this treatment was unacceptable. no data recorded for that.

my goal:

  ANOVA for testing main effects and interactions for significance. i mean, i want to test all treatments data with each other.

my GLM output: 1- a Tests of Between-Subjects Effects table. 2- Estimated Marginal Means table. 3-Post Hoc Tests for the factor with 4 levels.

my problem:

 the GLM gave me an output which there were no post-hoc test within groups. i can't find how to analyse the significance within each groups data and one data with other data in another group. does it need to run a separate one-Way Anova?