mdt: MDT

Description Usage Arguments Note

View source: R/mdt.R

Description

This function defines a melodic discrimination test (MDT) module for incorporation into a psychTestR timeline. Use this function if you want to include the MDT in a battery of other tests, or if you want to add custom psychTestR pages to your test timeline. For demoing the MDT, consider using demo_mdt(). For a standalone implementation of the MDT, consider using standalone_mdt().

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
mdt(
  num_items = 20L,
  take_training = TRUE,
  label = "MDT",
  feedback = mdt.feedback.no_score(),
  media_dir = "http://media.gold-msi.org/test_materials/MDT/v4",
  next_item.criterion = "bOpt",
  next_item.estimator = "BM",
  next_item.prior_dist = "norm",
  next_item.prior_par = c(0, 1),
  final_ability.estimator = "WL",
  constrain_answers = FALSE,
  dict = mdt::mdt_dict
)

Arguments

num_items

(Integer scalar) Number of items in the test.

take_training

(Logical scalar) Whether to include the training phase.

label

(Character scalar) Label to give the MDT results in the output file.

feedback

Defines the feedback to give the participant at the end of the test. By default no feedback is given. This can be a timeline segment (as created by new_timeline), a test element (as created by e.g. page), or a list of test elements. The following built-in choices are available see function-level documentation for details):

  • mdt.feedback.no_score

  • mdt.feedback.simple_score

  • cat.feedback.graph

  • cat.feedback.irt

  • cat.feedback.iq

media_dir

(Character scalar) File path to the directory hosting the test's media files (typically a publicly accessible web directory).

next_item.criterion

(Character scalar) Criterion for selecting successive items in the adaptive test. See the criterion argument in nextItem for possible values. "bOpt" corresponds to the setting used in the original MDT paper.

next_item.estimator

(Character scalar) Ability estimation method used for selecting successive items in the adaptive test. See the method argument in thetaEst for possible values. "BM", Bayes modal, corresponds to the setting used in the original MDT paper. "WL", weighted likelihood, corresponds to the default setting used in versions <= 1.3.0 of this package.

next_item.prior_dist

(Character scalar) The type of prior distribution to use when calculating ability estimates for item selection. Ignored if next_item.estimator is not a Bayesian method. Defaults to "norm" for a normal distribution. See the priorDist argument in thetaEst for possible values.

next_item.prior_par

(Numeric vector, length 2) Parameters for the prior distribution; see the priorPar argument in thetaEst for details. Ignored if next_item.estimator is not a Bayesian method. The dfeault is c(0, 1).

final_ability.estimator

Estimation method used for the final ability estimate. See the method argument in thetaEst for possible values. The default is "WL", weighted likelihood, which corresponds to the setting used in the original MDT paper. If a Bayesian method is chosen, its prior distribution will be defined by the next_item.prior_dist and next_item.prior_par arguments.

constrain_answers

(Logical scalar) If TRUE, then item selection will be constrained so that the correct answers are distributed as evenly as possible over the course of the test. We recommend leaving this option disabled.

dict

The psychTestR dictionary used for internationalisation.

Note

Versions <= 1.3.0 of this package experimented with weighted likelihood ability estimation for item selection. However, current versions of the package revert to Bayes modal ability estimation for item selection, for consistency with the original MDT paper.


pmcharrison/mdt documentation built on Sept. 16, 2020, 8:29 a.m.