Skip to main content
. Author manuscript; available in PMC: 2015 Mar 25.
Published in final edited form as: Clin Psychol Sci. 2014 Mar 1;2(2):187–201. doi: 10.1177/2167702613495200

Table 2.

Multilevel Models Predicting the Instability of Daily Affect and Self-Esteem

Outcome df β SE t-ratio p-value
Daily PA SSD
  Intercept 1, 77 −0.67 0.10 −6.86 < .001
  SAD 0.12 0.10 1.24 .221
  Intercept 4,74 −0.73 0.08 −8.67 < .001
  Age −0.03 0.01 −3.72 .001
  SAD 0.11 0.13 0.86 .391
  Mean PA −0.01 0.12 −0.07 .949
  SAD × Mean PA 0.35 0.12 3.00 .004
Daily NA SSD
  Intercept 1, 77 −0.85 0.14 −6.27 < .001
  SAD 0.40 0.14 2.91 .005
  Intercept 3, 75 −1.20 0.08 −15.44 < .001
  SAD 0.06 0.09 0.65 .516
  Mean NA 1.00 0.08 12.03 < .001
  SAD × Mean NA −0.38 0.08 −4.63 < .001
Daily Self-Esteem SSD
  Intercept 1, 76 0.48 0.09 5.40 < .001
  SAD 0.36 0.09 4.03 < .001
  Intercept 3, 74 0.43 0.09 4.90 < .001
  SAD 0.13 0.10 1.37 .174
  Mean Self-Esteem −0.49 0.13 −3.71 .001
  SAD × Mean Self-Esteem 0.23 0.13 1.76 .083

Notes.Group differences were tested by generalized linear models with log link. Only diagnostic group contrast was entered in first models for each construct. Gender, age, group contrast, mean levels, and interaction of group and mean level were initially entered in the second models; gender and/or age were removed if non-significant. NA = negative affect; PA = positive affect; SAD = Social Anxiety Disorder Diagnosis (contrast coded); SSD = squared successive differences.