Table 1. Fit Statistics for Guessing, Threshold, Associative Learning, and Hypothesis Testing Models of Concept Identification Learning.
Fit Statistics | |||||
---|---|---|---|---|---|
|
|||||
Model | Parameters | Deviance | pD | WAIC | K-S D Errors |
Guessing | 0 | 3,945,960.00 | -- | -- | 0.77 |
Threshold | 101,785 | 2,376,306.00 | 0.00 | ||
Associative Learning | |||||
Guessing and Learned | 35,556 | 1,862,407.00 | 24,816.66 | 1,923,774.00 | 0.13 |
Perseverative, Guessing, and Learned | 71,111 | 1,832,287.00 | 29,067.85 | 1,904,628.00 | 0.12 |
Hypothesis Testing | |||||
Guessing and Learned | 35,556 | 1,844,850.00 | 22,528.04 | 1,902,575.00 | 0.13 |
Perseverative, Guessing, and Learned | 71,111 | 1,803,620.00 | 28,024.10 | 1,876,088.00 | 0.12 |
Note. Parameters = total number of parameters; Deviance = -2log-likelihood; pD = effective number of parameters; WAIC = Wantanabe-Akaike information criterion; K-S D Errors= Kolmogorov-Smirnov distance comparing observed and model expected cumulative distributions of total errors. Guessing and Threshold model parameters were not estimated, and thus these have no pD or WAIC values.