Table 4.

Comparison of Bayesian Information Criterion (BIC), Weight of Evidence for BIC (bWeight), Akaike Information Criterion (AIC), and Weight of Evidence for AIC (aWeight) for Unadjusted Models of Iowa Gambling Task Performance Metrics

Overall Performance Model	BIC	bWeight	AIC	aWeight
1. Null model (intercept only)	6382.7	< .001	6324.8	< .001
2. Linear effect of age and random effect of intercept	6283.8	.51	6221.4	.01
3. Linear effect of age and random effects of intercept and age	6284.2	.43	6212.8	.82
4. Quadratic and linear effects of age and random effect of intercept	6287.9	.07	6221.0	.01
5. Quadratic and linear effects of age and random effects of intercept and age	6305.3	< .001	6216.1	.16
Decision-Making Under Ambiguity Model	BIC	bWeight	AIC	aWeight
1. Null model (intercept only)	5271.9	< .001	5213.9	< .001
2. Linear effect of age and random effect of intercept	5236.0	.18	5173.6	.002
3. Linear effect of age and random effects of intercept and age	5233.0	.81	5161.7	.93
4. Quadratic and linear effects of age and random effect of intercept	5242.2	.01	5175.3	.001
5. Quadratic and linear effects of age and random effect of intercept and age	5256.3	< .001	5167.1	.06
Decision-Making Under Risk Model	BIC	bWeight	AIC	aWeight
1. Null model (intercept only)	6008.5	< .001	5950.5	< .001
2. Linear effect of age and random effect of intercept	5899.0	.54	5836.6	.02
3. Linear effect of age and random effects of intercept and age	5906.5	.01	5835.2	.04
4. Quadratic and linear effects of age and random effect of intercept	5899.4	.45	5832.5	.16
5. Quadratic and linear effects of age and random effects of intercept and age	5918.5	< .001	5829.4	.77

Note. Bolded font indicates best-fitting model for the criterion. Likelihood ratio tests comparing linear and quadratic models with the same random effects (Model 2 vs. Model 4; Model 3 vs. Model 5) were significant (p ≤ .01). Bolded font indicates best-fitting model for the criterion