Table 4

Random-effects response models: Log odds ratios of response by attributes of the invited alter and the referrer

(1)(2)(3)(4)
Invited Alter's Survey Mode (ref.=in person) 
 Phone 0.71 1.30 0.71 1.29 
  (0.71) (0.81) (0.71) (0.81) 
 Web −1.29 −1.40 −1.29 −1.39 
    (0.76) (0.78) (0.76) (0.78) 

Invited Alter's Gender (ref.=male) 
 Female  0.05  0.05 
     (0.31)  (0.31) 

Invited Alter's Education (ref.=high school or less) 
 College  0.03  0.08 
   (0.86)  (0.89) 
 More than college  0.81  0.87 
   (0.87)  (0.90) 

Referrer's Gender (ref.=male) 
 Female   −0.21 −0.36 
      (0.31) (0.33) 

Referrer's Education (ref.=high school or less) 
 College   −0.31 −0.80 
    (0.54) (0.63) 
 More than college   −0.29 −1.25 
      (0.51) (0.65) 
Constant 2.27*** 2.03* 2.67*** 3.21*** 
    (0.39) (0.81) (0.64) (0.79) 
Referrer-Level Variance 13.62*** 17.35*** 13.60*** 17.16*** 
 (3.17) (4.90) (3.15) (4.78) 
Pseudo-R2 .32 .35 .32 .34 
BIC 3,660.60 3,375.34 3,680.79 3,393.63 
N 916 898 916 898 
(1)(2)(3)(4)
Invited Alter's Survey Mode (ref.=in person) 
 Phone 0.71 1.30 0.71 1.29 
  (0.71) (0.81) (0.71) (0.81) 
 Web −1.29 −1.40 −1.29 −1.39 
    (0.76) (0.78) (0.76) (0.78) 

Invited Alter's Gender (ref.=male) 
 Female  0.05  0.05 
     (0.31)  (0.31) 

Invited Alter's Education (ref.=high school or less) 
 College  0.03  0.08 
   (0.86)  (0.89) 
 More than college  0.81  0.87 
   (0.87)  (0.90) 

Referrer's Gender (ref.=male) 
 Female   −0.21 −0.36 
      (0.31) (0.33) 

Referrer's Education (ref.=high school or less) 
 College   −0.31 −0.80 
    (0.54) (0.63) 
 More than college   −0.29 −1.25 
      (0.51) (0.65) 
Constant 2.27*** 2.03* 2.67*** 3.21*** 
    (0.39) (0.81) (0.64) (0.79) 
Referrer-Level Variance 13.62*** 17.35*** 13.60*** 17.16*** 
 (3.17) (4.90) (3.15) (4.78) 
Pseudo-R2 .32 .35 .32 .34 
BIC 3,660.60 3,375.34 3,680.79 3,393.63 
N 916 898 916 898 

Notes: Standard errors are shown in parentheses. Pseudo-R2 is calculated using McFadden's pseudo-R2, which compares the given model to the null (intercept only) model. BIC=Bayesian information criterion.

p < .10; *p < .05; ***p < .001 (two-tailed tests)

Close Modal

or Create an Account

Close Modal
Close Modal