Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!
Get Started for Free

mle-methods

Compute maximum likelihood estimates of theta

Description

mle is a function to compute maximum likelihood estimates of theta.

Usage

mle(
  object,
  select = NULL,
  resp,
  start_theta = NULL,
  max_iter = 100,
  crit = 0.001,
  truncate = FALSE,
  theta_range = c(-4, 4),
  max_change = 1,
  use_step_size = FALSE,
  step_size = 0.5,
  do_Fisher = TRUE
)

## S4 method for signature 'item_pool'
mle(
  object,
  select = NULL,
  resp,
  start_theta = NULL,
  max_iter = 50,
  crit = 0.005,
  truncate = FALSE,
  theta_range = c(-4, 4),
  max_change = 1,
  use_step_size = FALSE,
  step_size = 0.5,
  do_Fisher = TRUE
)

MLE(
  object,
  select = NULL,
  start_theta = NULL,
  max_iter = 100,
  crit = 0.001,
  theta_range = c(-4, 4),
  truncate = FALSE,
  max_change = 1,
  do_Fisher = TRUE
)

## S4 method for signature 'test'
MLE(
  object,
  select = NULL,
  start_theta = NULL,
  max_iter = 100,
  crit = 0.001,
  theta_range = c(-4, 4),
  truncate = FALSE,
  max_change = 1,
  do_Fisher = TRUE
)

## S4 method for signature 'test_cluster'
MLE(object, select = NULL, start_theta = NULL, max_iter = 100, crit = 0.001)

Arguments

object

an item_pool object.
select

(optional) if item indices are supplied, only the specified items are used.
resp

item response on all (or selected) items in the object argument. Can be a vector, a matrix, or a data frame. length(resp) or ncol(resp) must be equal to the number of all (or selected) items.
start_theta

(optional) initial theta values. If not supplied, EAP estimates using uniform priors are used as initial values. Uniform priors are computed using the theta_range argument below, with increments of .1.
max_iter

maximum number of iterations. (default = 100)
crit

convergence criterion to use. (default = 0.001)
truncate

set TRUE to impose a bound using theta_range on the estimate. (default = FALSE)
theta_range

a range of theta values to bound the estimate. Only effective when truncate is TRUE. (default = c(-4, 4))
max_change

upper bound to impose on the absolute change in theta between iterations. Absolute changes exceeding this value will be capped to max_change. (default = 1.0)
use_step_size

set TRUE to use step_size. (default = FALSE)
step_size

upper bound to impose on the absolute change in initial theta and estimated theta. Absolute changes exceeding this value will be capped to step_size. (default = 0.5)
do_Fisher

set TRUE to use Fisher scoring instead of Newton-Raphson method. (default = TRUE)

Value

mle returns a list containing estimated values.

  • th theta value.

  • se standard error.

  • conv TRUE if estimation converged.

  • trunc TRUE if truncation was applied on th.

Examples

mle(itempool_fatigue, resp = resp_fatigue_data[10, ])
mle(itempool_fatigue, select = 1:20, resp = resp_fatigue_data[10, 1:20])
TestDesign

Optimal Test Design Approach to Fixed and Adaptive Test Construction

v1.2.2
GPL (>= 2)
Authors
Seung W. Choi [aut, cre] (<https://orcid.org/0000-0003-4777-5420>), Sangdon Lim [aut] (<https://orcid.org/0000-0002-2988-014X>)
Initial release
2021-01-26

We don't support your browser anymore

Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.