Skip to main content
. Author manuscript; available in PMC: 2022 May 8.
Published in final edited form as: Aggress Behav. 2016 Feb 26;42(6):542–554. doi: 10.1002/ab.21649

Table 2.

Summary of Two-Step Linear Regression Analyses Used to Examine Main and Interaction Effects in a Model Predicting Frequency of Past Aggression Toward Gay Men and Lesbians While Controlling for Sexual Prejudice.


Variable
b SE rpartial p
Step 1
  Age .000 .000 −.02 .770
  Education level .000 .000 −.11 .124
  Income .000 .000 .06 .406
  Sexual prejudice .15 .03 .31 .000
  AIDS-related stigma .08 .03 .20 .007
  Race (White) −.03 .02 −.14 .051
  Prior contact (No) −.02 .02 −.07 .366
Step 2
  Age .000 .000 −.02 .841
  Education level −.01 .000 −.12 .099
  Income .000 .000 .11 .126
  Sexual prejudice .15 .03 .31 .000
  AIDS-related stigma .29 .08 .27 .000
  Race (White) −.07 .04 −.13 .082
  Prior contact (No) −.04 .03 −.09 .237
  AR Stigma X Race −.21 .11 −.14 .053
  AR Stigma X Prior Contact −.27 .09 −.23 .002
  Race X Prior Contact .05 .04 .01 .179
  AR Stigma X Race X Prior Contact .27 .12 .16 .026

Note. N = 194. Values of 0.000 are less than 0.0005. AR Stigma = AIDS-related stigma. Prior Contact = prior contact with gay men and lesbians. Results of Model 2, which included all possible interaction terms involving the covariate sexual prejudice, were omitted because sexual prejudice was not a significant moderator. Whites were coded as the reference group for race. No, as opposed to yes, for prior contact with gay men and lesbians was coded as the reference group for prior contact. Thus, in regards to the present sample, this model refers to White men with no prior contact with gay men or lesbians.