Table 2.
Experiment 1: Best-fitting parameters for the EBRW model.
w 1 | .329 |
w 2 | .344 |
w 3 | [.326] |
M1 | 3.181 |
M2 | 1.415 |
M3 | 1.202 |
M4 | [1.000] |
θ 1 | 4.745 |
θ 2 | 1.361 |
θ 3 | 0.944 |
θ 4 | 0.711 |
PM | 1.053 |
Pθ | 1.470 |
u | 0.000 |
v | 0.377 |
OLD | 3.464 |
NEW | 3.513 |
μ | 261.441 |
κ | 55.391 |
Note: Parameter values in brackets are not free to vary. wj = attention-weight for dimension j; Mj = lag-j memory strength; θj = lag-j sensitivity; PM, Pθ = primacy multipliers on memory-strength and sensitivity; u, v = background-element strength constants; OLD, NEW = response threshold magnitudes; μ = mean residual RT; κ = random-walk time-scaling constant.