WARNING: 1. Because group sizes vary, the chi square difference tests involving diversity, contrast, and group effects are approximate.
GROUP ACTOR-PARTNER INTERDEPENDENCE MODEL
The group composition variable is Gender, a dichotomy consisting of Women and Men, and the outcome variable is Group Identification. The variable Gender is presumed to affect Group Identification in four different ways: the effect of the actor's own Gender or X, the effect of other group members' Gender or X', the effect of the similarity of the actor's Gender to the other group members' Gender or I, and the effect of the similarity of the others' Gender or I'. All of these variables are effect coded. For instance, X is coded Women = +1 and Men = -1. Any group which contains missing data on Gender is dropped from the analysis.
RESULTS
Descriptives
For the analysis, there are 263 persons in 58 groups. The group sizes range from 3 and 5. There are 2 all-Men groups and 14 all-Women groups. There are 95 (36.1%) Men and 168 (63.9%) Women. The means and standard deviations are presented in Table 1.
Complete Model
The effect of the actor's Gender is .004 (p = .958). Women score .008 units higher on Group Identification than Men. The effect of the Gender of the other group members' is -.189 (p = .181). An actor, all of whose other group members are Women, scores .379 units lower on Group Identification than a member all of whose other members are Men. The effect of the actor's similarity to others in the group on Gender is .248 (p = .050). An actor, who is totally similar to the other group members on Gender, scores .496 units higher on Group Identification than an actor who is totally dissimilar. The effect of the others' similarity in Gender is -.105 (p = .425). In a 3-person group, a person with others who are completely homogeneous, either all Women or Men, scores .210 units lower on Group Identification than a person with 1 Women and 1 Men others.
The variance due to groups is equal to .107 and the proportion of variance due to group or intraclass correlation is equal to .089. The multiple correlation explained by the four GAPIM-I fixed effects is equal to .090. The chi square test with four degrees of freedom that compares the Complete Model to the Empty Model equals 6.629 (p = .157). Because this chi square test is not statistically significant, we do not have evidence that any of the terms of the Complete Model are non-zero. The chi square test with two degrees of freedom that compares the Complete Model to the Main Effects Model equals 4.405 (p = .354). Because this chi square test is not statistically significant, we do not have evidence that the interaction effects of the Complete Model are non-zero. The sample size adjusted BIC for the Complete Model is 793.26 whereas the value for the Empty Model is 796.22. Because the index is smaller, the Complete Model is a better fitting model than the Empty Model.
The Best Fitting Model
The best fitting model is the Contrast (Main Effects and Interaction) Model with a multiple R of .105. This implies that there is a contrast effect both for the effect of Gender and the effects of similarity. The chi square test comparing this model to the Empty Model is equal to 5.159 with 2 degrees of freedom (p = .076). The sample size adjusted BIC is equal to 792.898.
Table 1: Descriptive Statistics
Variable Mean Standard Deviation
--------------------------------------------------------
Gender .278 .963
Group Identification 4.102 1.102
Table 2: Model Estimates and Fit
Model X X' I I' SABIC R
-----------------------------------------------------------------------------------------------------------
Empty 0 0 0 0 796.224
Main Effects .049 -.172 0 0 795.833 .029
Actor Only .038 0 0 0 796.872 .000
Others Only 0 -.162 0 0 795.375 .055
Group -.139 -.139 0 0 796.066 .019
Contrast .076 -.076 0 0 795.760 .036
Complete .004 -.189 .248* -.105 793.261 .090
Person Fit .003 -.235+ .243+ 0 793.012 .096
Diversity .039 -.205 .094 .094 796.456 .000
Contrast .018 -.138 .180+ -.180+ 793.015 .102
Constaints on both Main Effects and Interactions
Actor Only .002 0 .185 0 795.516 .000
Others Only 0 -.123 0 -.087 795.889 .000
Group -.189 -.189 .107 .107 796.606 .000
Contrast .045 -.045 .181+ -.181+ 792.898 .105
Actor and Others Similarity .046 0 0 -.146+ 796.318 .000
Others and Person Fit 0 -.123 .244* 0 792.097 .114
Table 3: Chi Square Tests
Model Tested Chi Square df p
------------------------------------------------------
EMPTY MODEL AS THE COMPARISON MODEL (significant means a better fitting model)
Main Effects 2.224 2 .329
Actor Only .269 1 .604
Others Only 1.766 1 .184
Group 1.075 1 .300
Contrast 1.381 1 .240
Complete 6.629 4 .157
Person Fit 5.962 3 .113
Diversity 2.517 3 .472
Contrast 5.959 3 .114
Constraints on both Main Effects and Interactions
Actor Only 2.541 2 .281
Others Only 2.168 2 .338
Group 1.451 2 .484
Contrast 5.159 2 .076
Actor and Others' Similarity 1.739 2 .419
Others and Person Fit 5.960 2 .051
MAIN EFFECTS MODEL AS THE COMPARISON MODEL (significant means a worse fitting model)
Actor Only 1.955 1 .162
Others Only .458 1 .498
Group 1.149 1 .284
Contrast .843 1 .359
COMPLETE MODEL AS THE COMPARISON MODEL (significant means a worse fitting model)
Person Fit .667 1 .414
Diversity 4.112 1 .043
Contrast .670 1 .413
Constraints on both Main Effects and Interactions
Actor Only 4.088 2 .130
Others Only 4.461 2 .107
Group 5.178 2 .075
Contrast 1.470 2 .480
Actor and Others' Similarity 4.890 2 .087
Others and Person Fit .669 2 .716