Adding two new models

Author: Rausch

.Rbuildignore

@@ -4,7 +4,6 @@ QuickTests.R
 usefultools.R
 ^.*\.Rproj$
 ^\.Rproj\.user$
-LICENSE
 README.rmd
 README_files
 paper.bib
@@ -17,3 +16,5 @@ StatConfR.pdf
 .github
 .Rd2pdf14268
 statConfR.Rcheck
+^dev$
+LICENSE

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: statConfR
 Type: Package
 Title: Models of Decision Confidence and Measures of Metacognition
-Version: 0.2.1
+Version: 0.2.2
 Date: 2025-09-24
 Authors@R: c(
   person("Manuel", "Rausch", email = "[email protected]", 

NAMESPACE

@@ -4,7 +4,6 @@ export(estimateMetaI)
 export(fitConf)
 export(fitConfModels)
 export(fitMetaDprime)
-export(groupBMS)
 export(plotConfModelFit)
 export(simConf)
 import(ggplot2)
@@ -20,6 +19,5 @@ importFrom(stats,optim)
 importFrom(stats,plnorm)
 importFrom(stats,pnorm)
 importFrom(stats,qnorm)
-importFrom(stats,rgamma)
 importFrom(stats,rmultinom)
 importFrom(utils,tail)

R/fitConf.R

@@ -20,7 +20,7 @@
 #'    a factor with a warning),
 #' * \code{correct} (encoding whether the response was correct; should  be 0 for incorrect responses and 1 for correct responses)
 #' @param model `character` of length 1. The generative model that should be
-#'    fitted. Models implemented so far: 'WEV', 'SDT', 'GN', 'PDA', 'IG',
+#'    fitted. Models implemented so far: 'WEV', 'SDT', 'GN', 'PDA', 'IG', 'RCE', 'CAS',
 #'    'ITGc', 'ITGcm', 'logN', and 'logWEV'.
 #' @param nInits `integer`. Number of starting values used for maximum likelihood optimization.
 #' Defaults to 5.
@@ -135,7 +135,7 @@
 #' amount of evidence available for the discrimination decision and  can be smaller
 #' as well as greater than 1.
 #'
-#' ### \strong{Logistic noise model (logN)}
+#' ### \strong{Lognormal noise model (logN)}
 #' According to logN, the same sample
 #' of sensory evidence is used to generate response and confidence, i.e.,
 #' \eqn{y=x} just as in SDT (Shekhar & Rahnev, 2021). However, according to logN, the confidence criteria
@@ -154,7 +154,7 @@
 #' \overline{\theta}_{-1,L-1}, \overline{\theta}_{1,1}, ...  \overline{\theta}_{1,L-1}},
 #' as free parameters.
 #'
-#' ### \strong{Logistic weighted evidence and visibility model (logWEV)}
+#' ### \strong{Lognormal weighted evidence and visibility model (logWEV)}
 #' logWEV is a combination of logN and WEV proposed by Shekhar and Rahnev (2023).
 #' Conceptually, logWEV assumes that the observer combines evidence about decision-relevant features
 #' of the stimulus with the strength of evidence about choice-irrelevant features (Rausch et al., 2018).
@@ -180,18 +180,19 @@
 #' The mean of \eqn{x_{-1}} is given by \eqn{(1 - S) \times 0.25 \times d}; the mean
 #' of \eqn{x_1} is given by \eqn{(1 + S) \times 0.25 \times d}. The sensory evidence
 #' used for the discrimination choice is \eqn{x = x_2 - x_1},
-#' which implies that the discrimination decision is equivalent to standard SDT.
+#' which implies that the process underlying the discrimination decision is equivalent to standard SDT.
 #' The confidence decision variable y is \eqn{y = - x_1} if the response R is -1 and \eqn{y = x_2} otherwise.
 #'
 #' ### \strong{CASANDRE (CAS)}
 #' Generation of the primary choice in the CASANDRE model follows standard SDT assumptions.
 #' For confidence, the CASANDRE model assumes an additional stage of processing based on the observer’s estimate of the
 #' perceived reliability of their choices (Boundy-Singer et al., 2023).
-#' The confidence variable \eqn(y) is equal to  \eqn{x / \hat{sigma}}. \eqn{\hat{sigma}}
-#' represents a noisy internal estimate of the sensory noise.
-#' It is assumed that \eqn{\hat{sigma}} is sampled from a lognormal distribution with a mean fixed to 1
-#' and a free noise parameter sigma. Conceptually, sigma represents the uncertainty
-#' in an individual's estimate of their own sensory uncertainty.
+#' The confidence decision variable y is given by \eqn{y = \frac{x}{\hat{\sigma}}}.
+#' \eqn{\hat{\sigma}} represents a noisy internal estimate of the sensory noise.
+#' It is assumed that \eqn{\hat{\sigma}} is sampled from a lognormal distribution with a mean fixed to 1
+#' and a free noise parameter \eqn{\sigma}.
+#' Conceptually, \eqn{\sigma} represents the uncertainty in an individual's estimate of their own sensory uncertainty.
+#'
 #'
 #' @author Sebastian Hellmann, \email{[email protected]}\cr
 #' Manuel Rausch, \email{[email protected]}

R/fitConfModels.R

@@ -30,7 +30,7 @@
 #'    the models given in the second argument are fitted to each subset of `data`
 #'    determined by the different values of this column)
 #' @param models `character`. The different computational models that should be
-#'    fitted. Models implemented so far: 'WEV', 'SDT', 'GN', 'PDA', 'IG', 'ITGc', 'RCE',
+#'    fitted. Models implemented so far: 'WEV', 'SDT', 'GN', 'PDA', 'IG', 'ITGc', 'RCE', 'CAS',
 #'    'ITGcm', 'logN', and 'logWEV'. Alternatively, if `model="all"` (default),
 #'    all implemented models will be fit.
 #' @param nInits `integer`. Number of initial values used for maximum likelihood optimization.
@@ -160,7 +160,7 @@
 #' amount of evidence available for the discrimination decision and  can be smaller
 #' as well as greater than 1.
 #'
-#' ### \strong{Logistic noise model (logN)}
+#' ### \strong{Lognormal noise model (logN)}
 #' According to logN, the same sample
 #' of sensory evidence is used to generate response and confidence, i.e.,
 #' \eqn{y=x} just as in SDT (Shekhar & Rahnev, 2021). However, according to logN, the confidence criteria
@@ -179,7 +179,7 @@
 #' \overline{\theta}_{-1,L-1}, \overline{\theta}_{1,1}, ...  \overline{\theta}_{1,L-1}},
 #' as free parameters.
 #'
-#' ### \strong{Logistic weighted evidence and visibility model (logWEV)}
+#' ### \strong{Lognormal weighted evidence and visibility model (logWEV)}
 #' logWEV is a combination of logN and WEV proposed by Shekhar and Rahnev (2023).
 #' Conceptually, logWEV assumes that the observer combines evidence about decision-relevant features
 #' of the stimulus with the strength of evidence about choice-irrelevant features (Rausch et al., 2018).
@@ -205,18 +205,18 @@
 #' The mean of \eqn{x_{-1}} is given by \eqn{(1 - S) \times 0.25 \times d}; the mean
 #' of \eqn{x_1} is given by \eqn{(1 + S) \times 0.25 \times d}. The sensory evidence
 #' used for the discrimination choice is \eqn{x = x_2 - x_1},
-#' which implies that the discrimination decision is equivalent to standard SDT.
+#' which implies that the process underlying the discrimination decision is equivalent to standard SDT.
 #' The confidence decision variable y is \eqn{y = - x_1} if the response R is -1 and \eqn{y = x_2} otherwise.
 #'
 #' ### \strong{CASANDRE (CAS)}
 #' Generation of the primary choice in the CASANDRE model follows standard SDT assumptions.
 #' For confidence, the CASANDRE model assumes an additional stage of processing based on the observer’s estimate of the
 #' perceived reliability of their choices (Boundy-Singer et al., 2023).
-#' The confidence variable \eqn(y) is equal to  \eqn{x / \hat{sigma}}. \eqn{\hat{sigma}}
-#' represents a noisy internal estimate of the sensory noise.
-#' It is assumed that \eqn{\hat{sigma}} is sampled from a lognormal distribution with a mean fixed to 1
-#' and a free noise parameter sigma. Conceptually, sigma represents the uncertainty
-#' in an individual's estimate of their own sensory uncertainty.
+#' The confidence decision variable y is given by \eqn{y = \frac{x}{\hat{\sigma}}}.
+#' \eqn{\hat{\sigma}} represents a noisy internal estimate of the sensory noise.
+#' It is assumed that \eqn{\hat{\sigma}} is sampled from a lognormal distribution with a mean fixed to 1
+#' and a free noise parameter \eqn{\sigma}.
+#' Conceptually, \eqn{\sigma} represents the uncertainty in an individual's estimate of their own sensory uncertainty.
 #'
 #' @author
 #' Sebastian Hellmann, \email{[email protected]}\cr

R/int_fitRCE.R

@@ -106,7 +106,6 @@ ll_RCE <-
     p <- c(t(p))
     ds <- cumsum(exp(p[1:(nCond)])) # enforce that sensitivity is ordered
 
-    b <- p[nCond+nRatings]
     theta <- p[nCond+nRatings]
     c_RA <- c(-Inf, p[nCond+nRatings-1] -
                 rev(cumsum(c(exp(p[(nCond+1):(nCond+nRatings-2)])))),
@@ -148,8 +147,9 @@ ll_RCE <-
 
     p_SB_RB <- outer(1:nCond, 1:nRatings, P_SBRB)
     p_SB_RA <- outer(1:nCond, 1:nRatings, P_SBRA)
-    p_SA_RA <- outer(1:nCond, 1:nRatings, P_SARA)
-    p_SA_RB <- outer(1:nCond, 1:nRatings, P_SARB)
+    p_SA_RA <- outer(1:nCond, 1:nRatings, P_SARA)  # rowSums(p_SB_RA) + rowSums(p_SB_RB)
+    p_SA_RB <- outer(1:nCond, 1:nRatings, P_SARB) # rowSums(p_SA_RA) + rowSums(p_SA_RB)
+
 
     p_SB_RB[(is.na(p_SB_RB))| is.nan(p_SB_RB)| p_SB_RB < 10^-64] <- 10^-64
     p_SB_RA[(is.na(p_SB_RA))| is.nan(p_SB_RA)| p_SB_RA < 10^-64] <- 10^-64

R/plotConfModelFit.R

@@ -120,8 +120,8 @@ plotConfModelFit <- function(data, fitted_pars, model = NULL){
            'IG' = "Independent Gaussian model",
            'ITGc'  = "Independent truncated Gaussian model: HMetad-Version",
            'ITGcm' = "Independent truncated Gaussian model: Meta-d'-Version",
-           'logN' = "Logistic noise model",
-           'logWEV' = "Logistic weighted evidence and visibility model",
+           'logN' = "Lognormal noise model",
+           'logWEV' = "Lognormal weighted evidence and visibility model",
            'PDA' = "Post-decisional accumulation model",
            'WEV' = "Weighted evidence and visibility model",
            'SDT' = "Signal detection rating model",