Skip to main content
. 2022 Oct 12;11(4):1002–1011. doi: 10.1556/2006.2022.00073

Table 2.

Summary of GLME models predicting probability of reporting harm within each domain as a function of time (gamblers only)

Link function: Binomial (logit)
Dependent variable: p(harm)
Financial Relat. Emotional/Psych. Health Work/Stud. Other
Fixed effects B (SE)
log (time +1) −0.862* −0.962* −0.969* −0.779* −0.797 −0.839*
(0.092) (0.149) (0.126) (0.167) (0.342) (0.323)
Constant −2.953* −3.966* −2.882* −5.544* −9.046* −8.535*
(0.356) (0.306) (0.255) (0.700) (0.716) (0.650)
Random effects SD
Response IDa 1.584 2.481 2.189 3.222 6.246 5.790
Harm IDb 1.302 0.305 0.514 0.537 0.739 0.565
Observations 10,448 8,489 7,183 10,448 7,836 10,448
Log Likelihood −2,128.14 −1,305.03 −1,603.56 −1,318.07 −533.36 −658.40
Akaike Inf. Crit. 4,264.28 2,618.07 3,215.13 2,644.13 1,074.72 1,324.80
Bayes. Inf. Crit. 4,293.29 2,646.25 3,242.64 2,673.15 1,102.58 1,353.81

Note: *P < 0.01, a The standard deviation of random intercepts across individuals, b The standard deviation of random intercepts across specific harms within each domain.