# threshold: Threshold Model In JanaJarecki/cogscimodels: Cognitive Models - Estimation, Prediction, and Development of Models for Cognitive Scientists

## Description

`treshold()` fits a threshold model.

## Usage

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30``` ```threshold( formula, data, fix = list(), choicerule = NULL, mode, discount = 0L, options = list(), ... ) threshold_c( formula, data, fix = list(), choicerule = NULL, discount = 0, options = list(), ... ) threshold_d( formula, data, fix = list(), choicerule = "softmax", discount = 0, options = list(), ... ) ```

## Arguments

 `formula` A formula specifying choice ~ var `...` other arguments from other functions, currently ignored.

## Details

• `treshold_c()`, which is the continuous model, fits the numeric distance to a threshold

• `treshold_d()`, which is the discrete model, fits choices given the distance to a threshold

Given the formula `y ~ a` the model predicts y = 1 for a >= `nu` and y = 0 for a < `nu`

## Value

A model of class "treshold". It can be viewed with `summary(mod)`, where `mod` is the name of the model object.

## Parameter Space

 ` `Name ` `LB - UB` ` Description Start Value ` ``nu` -Inf - Inf Treshold 0

Jana B. Jarecki

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32``` ```D <- data.frame( y = rep(0:1, each=5), a = 1:10) M <- threshold_c(y ~ a, D, fix="start") # fixed par. to start values predict(M) # predict dist. to threshold anova(M) # anova-like table summary(M) # summarize M <- threshold_d(y ~ a, D, fix="start") # fixed par. to start values predict(M) # predict dist. to threshold anova(M) # anova-like table summary(M) # summarize M\$MSE() # mean-squared error ### Binary response given a threshold # -------------------------------------------- M <- threshold(y ~ a, D, fix="start", choicerule = "softmax") predict(M) # --"-- maximum posterior anova(M) # anova-like table summary(M) # summarize M\$MSE() # mean-squared error ### Parameter specification and fitting ---------------------------------------- # Use a response variable, y, to which we fit parameter threshold(y ~ a, D, fix = "start", "softmax") # "start" fixes all par., # and fits none threshold(y ~ a , D, list(nu=2), "softmax") # fix threshold nu to 2 threshold(y ~ a, D, list(tau=0.5), "softmax") # fix soft-max tau to 1 threshold(y ~ a, D, choicerule = "softmax") # nu and tau free param ```

JanaJarecki/cogscimodels documentation built on Sept. 8, 2020, 7:28 p.m.