R: Generate full IRT report based on 'mirt' package

irt.report {dodoR}

R Documentation

Generate full IRT report based on `mirt` package

Description

This function generates a full report of an IRT run of the mirt package. It plots all possible charts for the test and each of the items; saves item parameters and item fit statistics (like Zhs and X2) into csv files; and uses a knitr template to generate a PDF report.

Usage

irt.report(answers, out, keys = NULL, author = "",
  title = "dodoR - Relatório", out.stats = NULL, out.itemplots = NULL,
  out.testplots = NULL, method = "EM", optimizer = "NR", verbose = F,
  test_score = T, test_info = T, test_SE = T, test_infoSE = T,
  trace = T, info = T, se = T, score = T, infoSE = T, infotrace = T)

Arguments

`answers`	A `data.frame` containing examinees answers to items. Rows must represent examinees and columns represent items. If the `data.frame` is in binary format, `keys` is not required, but distractor analysis is not perfoermed.
`out`	A directory to which all files will be saved. If the directory does not exist, it will be recursively created.
`keys`	A vector containing the keys of the test, that is, the right answers for each item in the correct order.
`author`	The author of the document (appears at the report cover)
`title`	The title of the document (appears at the report cover)
`out.stats`	The directory to which CSV files will be saved. If the directory does not exist, it will be recursively created. if out.stats = NULL, it will be a subfolder of out.
`out.itemplots`	The directory to which item plots will be saved. If the directory does not exist, it will be recursively created. if out.itemplots = NULL, it will be a subfolder of out.
`out.testplots`	The directory to which test plots will be saved. If the directory does not exist, it will be recursively created. if out.testplots = NULL, it will be a subfolder of out.
`method`	Estimation method. This parameter is passed directly to `mirt`. Default is 'EM', the Expectation-Maximization algorithm. For more options, consult the `mirt` manual.
`optimizer`	Numerical approximation method used for optimizing convergence of estimation. This parameter is passed directly to `mirt`. Default is 'NR', the Newton-Raphson method. For more options, consult the `mirt` manual.
`verbose`	Defines how much output the algorithm must give. This parameter is passed directly to `mirt`. Default is FALSE. For more options, consult the `mirt` manual.
`test_score`	Whether or not test score plot should be generated.
`test_info`	Whether or not test information plot should be generated.
`test_SE`	Whether or not test standard error plot should be generated.
`test_infoSE`	Whether or not a plot containing both test information and standard error should be generated.
`trace`	Whether or not item trace plots should be generated.
`info`	Whether or not item information plots should be generated.
`se`	Whether or not item standard error plots should be generated.
`score`	Whether or not item score plots should be generated.
`infoSE`	Whether or not a plot containing both item information and standard error should be generated.
`infotrace`	Whether or not a plot containing both item information and trace should be generated.

Value

function returns an object of class SingleGroupClass (SingleGroupClass-class), the same one returned by the 'mirt-package' function 'mirt', in case additional processing is desired beyond the generation of plots and CSV files with parameters

Author(s)

Douglas De Rizzo Meneghetti