# parameter recovery - no-pooling approach
- results look better than [[220330_203004 parameter recovery - partial-pooling with stan|partial-pooling with stan]]
- simulation parameters
- 400 subjects, 18 trials each
- fit/recover parameters with `optim` (5 fits/attempts per subject to try to find global minimum)
- spearman's rho (ranking)
- blue: correlation
- red: diagonal (dots on this line indicate perfect fit)
- x: true parameter values
- y: recovered parameter values
![[parameter_recovery.png]]
# simulate crt correlations with parameters
- left panel
- x: true parameter
- y: crt
- right panel
- x: recovered parameter
- y: crt
- top/bottom: spearman's rho vs. pearson's r
## alpha
![[s20220330_195345.png]]
![[s20220330_195730.png]]
## tau
![[s20220330_195501.png]]
![[s20220330_195810.png]]
## beta
![[s20220330_195534.png]]
![[s20220330_195855.png]]
## delta
![[s20220330_195551.png]]
![[s20220330_195918.png]]