I've just been learning about the Expectancy Valence Model of the Iowa Gambling Task (see Busemeyer & Stout, 2002; Yechiam et al 2005). The model includes three parameters: motivation, learning-rate, and response sensitivity. The parameters aim to assess component processes when performing the task. Alternatively, an overall performance measure can be obtained such as wins minus losses.
The hope of model fitting in these cases is to better understand the causes of particular performance deficits. Such deficits might then be used to identify general decision making deficits in the individual.
More broadly, I am interested in the general psychometric properties of parameters obtained from mathematical models of task performance. I'm also interested in the interface between cognitive modelling, measuring individual differences, and imbuing model parameters with psychological import. Thus, with a normal psychological test we would be interested in reliability (e.g., test-retest reliability over various time frames) and correlates with related measures.
Thus, my main question:
What is the test-retest reliability of IOWA gambling task performance and parameters of the Expectancy Valence Model?
I'd also be interested in learning about how such reliability estimates vary with properties of the task. For example, increasing the number of trials presumably increases reliability.
- Busemeyer, J.R., & Stout, J.C. (2002). A contribution of cognitive decision models to clinical assessment: Decomposing performance on the Bechara gambling task. Psychological Assessment, 14, 253– 262.
- Yechiam, E., Busemeyer, J. R., Stout, J. C., & Bechara, A. (2005). Using cognitive models to map relations between neuropsychological disorders and human decision-making deficits. Psychological Science, 16(12), 973-978. PDF