- Added
kappa_coef()
function to calculate Cohen's weighted/unweighted Kappa coefficient. Response
objects prints 'misc' fields without an error when there is a factor type variable.- Fixed an error in the function
irt:::marginal_reliability(..., method = "empirical")
.
- Plot functions can now suppress plot titles by setting
title = NULL
. item_analysis()
function now has a new argumentstats
where only selected item analysis statistics can be calculated.item_analysis()
function can now calculatepval_unadj
which is simply the mean of item scores.pval
is still the mean of item scores divided by each item's maximum possible score.- Added
area_between_icc()
function to calculate the area between two item characteristic curves of two unidimensional dichotomous items. The function can calculate either exact area (except when 3PL with different pseudo-guessing parameters) or area between two theta values. cat_sim()
function can use MAP (maximum-a-posteriori or Bayes Modal) and MAP + ML (i.e. MAP until an imperfect response string and ML after) ability estimation methods.- Responses, items administered, administered item IDs, estimates before or
after the administration of an item can be extracted from the individual
CAT output. For details, see "?
$.cat_output
". - Added
output_type
argument toest_ability(, output_type = )
function with available following output types:"list"
,"data.frame"
,"tibble"
. Previously, the output was only a "list".
- Bug fix: An
Itempool
object with one testlet can be printed without an error. - Added
$item_ids
to testlet object to get the ids of the items within the testlet. - Fixed bugs in
plot_empirical_icc()
function.
- Added
$max_score
forTestlet
objects. For a givenTestlet
objectt1
,t1$max_score
returns the maximum obtainable score by all items within the testlet. - Added
$resp_item_list
forItempool
objects. For anItempool
objectip
,ip$resp_item_list
will combine items that are not in a testlet and items within a testlet and returns a list object that is composed of onlyItem
objects.
- Package can be used in MacOS as well. Resolved compiler related issues.
- Added
est_bilog()
function to estimate item parameters for "1PL", "2PL" and "3PL" models using BILOG-MG on computers with Windows OS. This function requires an installed BILOG-MG program on the machine.
- This is the initial release of the package.