Skip to main content
. Author manuscript; available in PMC: 2014 Jul 22.
Published in final edited form as: Psychol Rev. 2013 Jul;120(3):697–719. doi: 10.1037/a0033152

Table 1.

Model Parameters

Model and parameters Parameter symbol
Ter st a σ sb
Confidence model 384 76 0.038 0.093 0.78
Three-choice model 392 155 0.066 0.120 1.04

Decision boundaries b1 b2 b3 b4 b5 b6
Confidence model 1.64 1.71 2.41 2.53 1.92 1.75
Three-choice model 1.46 1.48 2.26

Confidence criteria c1 c2 c3 c4 c5
Confidence model −0.96 −0.06 0.67 1.56 2.44

Memory strength means μhfN μhfS μhfW μlfN μlfS μlfW
Confidence model 0.00 2.20 1.81 0.68 3.04 2.37

Memory strength standard deviations shfN shfS shfW slfN slfS slfW
Confidence model 0.85 1.19 1.29 0.92 1.58 1.90

Three-choice model drift rates v1 v2 v3
40:10:10 0.56 0.20 0.24
30:20:10/40:30:10 0.46 0.34 0.20
10:10:30 0.20 0.21 0.59

Note. Ter is the mean nondecision time, st is the range in nondecision time, a is the scaling factor that multiplies drift rate, σ is the standard deviation in within trial variability, sb is the range in variability in the decision boundaries, and b1b6 are decision bounds. For the three-choice model, b3 is higher than b1 and b2 because it is correct for a low proportion of the trials. For the confidence model, c1c5 are confidence criteria from the division between high confidence new and medium confidence new to the division between medium confidence old and high confidence old, hf represents high frequency, lf represents low frequency, N represents new, W represents weak studied items (1.5 s per pair), and S represents strong studied items (5 s per pair). For the three-choice model, across trial standard deviation in drift rate for the largest drift rate was .042.