| EZ2 {EZ2} | R Documentation |
Fit the a simplified diffusion model for response time and accuracy to observed proportions of errors and response time means and variances.
EZ2(pstart, ObsValPair, ..., method = "Nelder-Mead", control = list(), hessian = FALSE)
pstart |
Vector with named starting parameter values |
ObsValPair |
Observed-predicted value pair in the form 0.80 ~ EZ2.pe(v1, z, a) or vrt2 ~ EZ2.pe(v2, a-z, a) if vrt2 exists in the global environment and is numeric |
... |
More Observed-predicted value pairs (as many as desired, see example) |
method |
See method parameter of optim. |
control |
See control parameter of optim. |
hessian |
See hessian parameter of optim. |
EZ2 fits a simplified version of the diffusion model for human and monkey response times and accuracy to the means and variances of the observables. This model of information accumulation and decision is a simplified version of Ratcliff's diffusion model (1978).
Use EZ2batch for more user friendly fitting automatically each row in a data.frame.
The list returned by optim. This list contains
par |
A vector containing the estimates |
value |
Sum of squared prediction errors. This should be very close to zero (order of 1e-8) if there are as many observed-predicted moment value pairs as there are unknown parameters (the estimates then constitute method of moments estimators). |
For other list members, see Value section of optim for a complete description.
Raoul Grasman
Ratcliff. Theory of Memory Retrieval. Psychological review (1978) vol. 85 (2) pp. 59-108
Grasman et al. EZ2: An extension of teh EZ-diffusion model for Response Time and Accuracy. Journal of Mathematical Psychology (2007) submitted
Vrt0 <- EZ2.vrt(0.1, 0.08, 0.12) Pe0 <- EZ2.pe(0.1, 0.08, 0.12) Vrt1 <- EZ2.vrt(0.15, 0.12-0.08, 0.12) Pe1 <- EZ2.pe(0.15, 0.12-0.08, 0.12) EZ2(c(v0=.156,v1=.0844,z=.0631,a=.1263), Vrt0 ~ EZ2.vrt(v0,z,a), Pe0 ~ EZ2.pe(v0,z,a), Vrt1 ~ EZ2.vrt(v1,a-z,a), Pe1 ~ EZ2.pe(v1, a-z, a))