Skip to main content

Table 2 Model summaries and beta coefficients for propensity and causal models of health utility scores

From: Benchmarking gambling screens to health-state utility: the PGSI and the SGHS estimate similar levels of population gambling-harm

   Beta coefficients (SE)
Model   Propensity Causal Causal (no covariates)
DV   SGHS
(0 vs 1+)
PGSI
(0 vs 1+)
SF-6D SF-6D SF-6D SF-6D
Regression type   Logistic Logistic OLS OLS OLS OLS
DV   SGHS PGSI SGHS PGSI SGHS PGSI
Constant   2.042***
(0.257)
2.435***
(0.235)
0.836***
(0.009)
0.832***
(0.009)
0.803***
(0.004)
0.804***
(0.004)
Gambling harms (0) None (n = 1546)     
(SGHS) (1-2) Low (n = 370)    −0.020**
(0.006)
  −0.022**
(0.007)
 
(3-5) Moderate (n = 368)    −0.062***
(0.007)
  −0.075***
(0.007)
 
(6-10) High (n = 319)    −0.109***
(0.007)
  −0.153***
(0.008)
 
Gambling problems (0) Non-problem NP (n = 1331)     
(PGSI) (1,2) Low risk LR (n = 399)     −0.005
(0.006)
  −0.007
(0.007)
(3-7) Moderate risk MR (n = 438)     −0.051***
(0.006)
  − 0.066***
(0.007)
(8+) Problems PG (n = 435)     −0.099***
(0.007)
  −0.137***
(0.007)
Alcohol consumption Non-drinker     
(AUDIT-C) (0-3) Non-risky    0.008*
(0.007)
0.011*
(0.008)
  
(4+) Risky    0.007
(0.007)
0.011
(0.007)
  
Age (polynomial) Linear (1) −0.034***
(0.003)
−0.037***
(0.003)
0.278*
(0.124)
0.111
(0.123)
  
Quadratic (2)    −0.251
(0.118)
−0.365**
(0.116)
  
Cubic (3)    −0.235
(0.116)
−0.246*
(0.113)
  
Gender Male   
Female −0.160
(0.107)
−0.251
(0.090)
− 0.016***
(0.005)
−0.019***
(0.005)
  
Country of birth Overseas   
Australia −0.164
(0.107)
−0.264*
(0.105)
− 0.015*
(0.006)
−0.008
(0.006)
  
Education Secondary or less     
Trade/Cert −0.091
(0.114)
−0.227*
(0.111)
    
Tertiary 0.212
(0.120)
−0.110
(0.119)
    
Postgrad 0.344*
(0.153)
0.200
(0.154)
    
Unemployed (ref = no)    −0.284
(0.195)
    
Personal income   0.111***
(0.029)
0.079**
(0.029)
    
Household income   −0.128***
(0.025)
−0.095***
(0.024)
    
Mother’s highest education achieved   −0.063***
(0.019)
     
Sick or on a disability pension (ref = no)     −0.126***
(0.015)
−0.133***
(0.015)
  
Recreational drug use (ref = no)     −0.020*
(0.008)
−0.015
(0.008)
  
Cigarettes consumed per day Non-smoker (0)     
<  10    −0.014*
(0.007)
−0.019**
(0.007)
  
10+    −0.010
(0.007)
−0.018*
(0.007)
  
Past year diagnosis of … (ref = no) Mood disorder    −0.049***
(0.009)
−0.053***
(0.009)
  
Anxiety disorder    −0.066***
(0.007)
−0.071***
(0.007)
  
Personality disorder    −0.017
(0.013)
−0.023
(0.013)
  
Any other psych. Disorder    −0.037**
(0.012)
−0.027*
(0.012)
  
Observations 2603 2603 2603 2603 2603 2603
Adjusted R2    0.292 0.288 0.148 0.132
Residual Std. Error    0.165 0.163 0.181 0.181
Model df    18, 2584 18, 2584 3, 2599 32,599
F    60.693*** 59.427*** 151.3*** 132.9***
  1. Note: Propensity models are unweighted. Case weights for the causal models (both with and without covariates) calculated from estimated probabilities from the propensity models. * p < 0.05, p < 0.01, p < 0.001