Skip to main content
. Author manuscript; available in PMC: 2021 Dec 13.
Published in final edited form as: Behav Ther. 2014 Sep 18;46(2):257–271. doi: 10.1016/j.beth.2014.09.006

Table 4.

Separate Generalized Estimating Equations Analyses Using a Gamma Distribution of the Lagged Effects of Body-, Eating-, and Exercise-Related Social Comparison and Body Surveillance on Body Dissatisfaction Controlling for Negative Affect

For the model involving the predictor variable of body-related social comparison:
Parameter Estimates B SE Wald χ2 df p
Outcome variable: body dissatisfaction, NLevel 1 = 8742, NLevel 2 = 232
Intercept 2.74 .17 277.97 1 <.001
Lagged Body-related Social Comparison (person-mean centered) .002 .001 10.96 1 .001
Body-related Social Comparison (person means) .04 .003 142.66 1 <.001
Time .001 .0001 22.41 1 <.001
Negative Affect (person-mean centered) .004 .01 .67 1 .414
Negative Affect (person means) −.02 .02 1.52 1 .218
(Scale) .91
Pseudo-R2 = .183
For the model involving the predictor variable of eating-related social comparison:
Parameter Estimates B SE Wald χ2 df p
Outcome variable: body dissatisfaction, NLevel 1 = 8742, NLevel 2 = 232
Intercept 3.11 .17 357.06 1 <.001
Lagged Eating-related Social Comparison (person-mean centered) .002 .001 16.32 1 <.001
Eating-related Social Comparison (person means) .03 .003 106.28 1 <.001
Time .001 .0001 18.17 1 <.001
Negative Affect (person-mean centered) .004 .01 .99 1 .319
Negative Affect (person means) −.02 .02 1.34 1 .248
(Scale) 1.00
Pseudo-R2 = .120
For the model involving the predictor variable of exercise-related social comparison:
Parameter Estimates B SE Wald χ2 df p
Outcome variable: body dissatisfaction, NLevel 1 = 8742, NLevel 2 = 232
Intercept 2.99 .18 288.26 1 <.001
Lagged Exercise-related Social Comparison (person-mean centered) .001 .001 4.65 1 .031
Exercise-related Social Comparison (person means) .03 .003 88.33 1 <.001
Time .001 .0001 18.82 1 <.001
Negative Affect (person-mean centered) .01 .01 .96 1 .326
Negative Affect (person means) .01 .02 .10 1 .748
(Scale) .81
Pseudo-R2 = .093
For the model involving the predictor variable of body surveillance:
Parameter Estimates B SE Wald χ2 df p
Outcome variable: body dissatisfaction, NLevel 1 = 8742, NLevel 2 = 232
Intercept 1.39 .24 34.74 1 <.001
Lagged Body Surveillance (person-mean centered) .03 .01 6.41 1 .011
Body Surveillance (person means) .42 .04 89.19 1 <.001
Time .001 .0001 23.89 1 <.001
Negative Affect (person-mean centered) .01 .004 1.45 1 .229
Negative Affect (person means) .03 .02 3.55 1 .060
(Scale) .78
Pseudo-R2 = .123