THIS IS A CRAN VERSION
IRT model/Dichotous models
section of the app (Thanks to Samsul Hadi for
reporting this issue).CLoSEread6
was added and it is also offered as a toy dataset
in the Data/Data
section of the app.DIF/Fairness
section of the app.plot_corr()
function is now more verbose and handles errors in polychoric
correlations estimation in a more user-friendly way; the partial matching of
arguments is no longer supportedData/Data
section of the app. Data/Data exploration
section of the app.DIF
section
of the app.Scores/Total scores
and DIF/Observed scores
section
of the app.Validity/Factor analysis
section of the app.Validity/Factor analysis
section of the app (Thanks to Stanislav
Ježek for suggesting this feature).HCIlong
dataset now contains zscore
variable.HeightInventory
dataset variables were renamed, new variable
total
was added.psychometric
package was removed.Anxiety
dataset is now offered as a toy dataset in the Data/Data
section
of the app.CZmaturaS
is newly offered as a toy dataset in the Data/Data
section of the app.THIS IS A CRAN VERSION
HCIlong
, MSATB
, GMAT
, EPIA
added anxiety
dataset was renamed to Anxiety
, variables score
and
zscore
were added total
was added to datasets HCI
, HCItest
Validity / Factor analysis
tab THIS IS A CRAN VERSION
gDiscrim()
function was fixed to assure indices are inside the defined
bounds, also for small samplesplotDistractorAnalysis()
function was fixed not to display double legends.
Thanks to Scarlett Escudero for reporting this issue.ItemAnalysis()
has been refactored substantially; memory demands are
lowered multiple times, as well as run times (most noticeable with large datasets).
The changes include following:NA
s for "gULI" when k
, l
, and u
arguments
are not specified, as there are no sensible defaults. Previous defaults
resulted in exactly the same values that ULI already provided on its own.Index.rel.drop
is not provided anymore, as it is not well defined
(see Gulliksen, 1950, for more details on reliability index (Index.rel
)).bin
and add.bin
arguments were removed, the analysis on binarized
data can be conducted by simply providing the cutscore in the cutscore
argument.criterion
argument now defaults to NULL
instead of "none"
, but
behaves the same.data
, y
are now defunct.DistractorAnalysis()
and plotDistractorAnalysis()
functions have been
improved: argument items allows options "all"
to apply for all items (default),
or a vector of item names (column names of Data
), or item identifiers
(integers specifying the column number).Anxiety
dataset was added.coef()
method as well.gDiscrim()
newly issues a warning when item(s) range is zero and when there
is no difficulty difference in either group.gDiscrim()
, the groups are newly defined using the quantile()
function.
In the case of ties, the respondents with the same total score are now in
the same group, the groups sizes may thus differ, but are now exactly defined.
The function may however provide different numerical results than in the previous
version.gDiscrim()
, you can now change how the dataset is split by passing
arguments to findInterval()
.HCI
data now includes total score variable. THIS IS A CRAN VERSION
It includes version 1.3.8 and following further changes:
Suggests
, meaning you are now not obliged to install any
of these dependencies if you plan to only use the package's "core" functions,
such as DDplot()
or ItemAnalysis()
.startShinyItemAnalysis()
function now checks for dependencies required
for its proper run and offers to install these dependencies if any of them
is not available.HeightInventory
was added.Validity/Factor analysis
section of the app, preventing crashes of the app
(Thanks to @koenderks for reporting this as issue #460)startShinyItemAnalysis()
gained an alias run_app()
.AttitudesExpulsion
was added.DIF-C
SIAmodule newly available from the DIF/Fairness
section of the
interactive appIRR-restricted
SIAmodule replaced the Reliability/IRR-restricted
section
of the interactive appValidity/Correlation structure
section.www
folder was cleaned up, .png replaced by .svg files improving the
resolution of plotsinst
folder structure was simplifiedTHIS IS A CRAN VERSION
MSclinical
, BFI2
, and TestAnxietyCor
were added, some
datasets and manuals were updated.fa_parallel()
was updated, smaller changes were made in some
other functionsValidity
section of the app, tab Factor analysis
was updated. IRT
section of the app, tab for the Bock model was updated to offer
parametrization closer to regression models.IRT
section.THIS IS A CRAN VERSION
Reports
section of the app the default name of the dataset was set
to the name of the csv file. This fixed some of the bugs with the HTML
report download.fa_parallel()
was added offering fast paralel analysis for
deciding on number of factors in exploratory factor analysis.CZmatura
and CZmaturaS
were added.Validity
section of the app, a new tab Factor analysis
was
added offering exploratory factor analysis.IRT models
section of the app, the binary IRT models were
collapsed into a single Dichotomous models
tab. IRT models
section of the app, the Bock IRT model section was
edited. DIF/Fairness
section of the app, the logistic and the polytomous
regression models now offer both the IRT and the intercept-slope
parametrization; the total score is newly offered as the matching criterion
besides the standardized total score.Criterion validity
and Item criterion validity
, the Pearson
correlation is newly used as the default option.Item analysis
section. Item analysis
section of the app, default discrimination was
changed to RIT. Observed ability
is now consistently used instead
of (DIF/DDF) Matching variable
. The changes done were especially in the
Data
, Regression models
and the DIF/Fairness
sections of the app, and
in the plotDIFLogistic()
function.Summary
of the app was renamed to Scores
to beter reflect focus
of this section. Validity
section of the shiny app, the dendrogram is newly enabled
even if numclust == 0.Item Analysis
section of the app, the tab Distractors plot
was
renamed to Item response curves
to better reflect that empirical response
curves are also available for binary and ordinal items.THIS IS A CRAN VERSION
DIF/About
section of the app were fixed (uniform vs.
nonuniform DDF is now correctly displayed).ItemAnalysis()
function was fixed.Regression
section of the app, missplaced SE
for intercepts and
slopes were fixed.Reliability/Restricted range
was added to the app offering
interactive examples demonstrating issues with reliability estimates in
restricted-range samples.ICCrestricted()
for computing ICC in restricted samples was
added.gDiscrim()
, DDplot()
and ItemAnalysis()
were updated,
variable names were unified.plot_corr()
to compute and visualize an item correlation
matrix was added and implemented within the ShinyItemAnalysis interactive
app.AIBS
, NIH
, HCIgrads
, HCItestretest
, HCIprepost
and HCIdata
were added.startShinyItemAnalysis()
function was fixed to finish job persisting in
the backgroud after the app is closed.DDF
section of the app.Regression
, IRT
and DIF/Fairness
sections of the app, with p
being the index for person and i
being the
index for an item.DIF/Fairness
section,
offering interpretation of group-specific models.Validity
, Regression
, IRT
, and DIF/Fairness
section
are now displayed in plotly mode in the shiny app.IRT/Rasch
section was restructured and it now offers both the
intercept/slope as well as the IRT parametrization. (This is planned for
other IRT
subsections in future versions.)Regression/Multinomial
section newly offers also the IRT
parametrization. DIF/Fairness
section.IRT/training
section.ggWright()
plot was updated to more general description.THIS IS A CRAN VERSION
Summary
and Item
subtabs
for the Generalized logistic regression method in the DIF/Fairness
section
of the app. Regression/Model comparison
section was
fixed.startShinyItemAnalysis()
function was updated to allow starting the
application as a background job in Rstudio (default setting).
This permits the user to work and try sample R code examples in the console.DIF/Fairness
sections for ordinal DIF.recode_nr()
for recognizing and recoding not-reached
items was introduced.ItemAnalysis()
function was updated with new columns for ratio of missed
and unreached items, shortened column names and ordering.Item analysis
section of the app
has been expanded to include item reliability, item validity, and percentage
of missing and unreached for each item.DDplot()
function now offers argument criterion
. Without criterion
specification, the difficulty-discrimination plot is provided. When
criterion
is specified, the difficulty-validity plot is provided instead,
depicting item criterion validity.Validity/Criterion_Validity/Items
section.DDplot()
was updated.THIS IS A CRAN VERSION
plotDIFLogistic()
now correctly plots matching criterion when item
purification is applied.DIF/Fairness
sections.DIF/Fairness
sections.DIF/Fairness
sections of shiny app.DIF/Fairness
section of the
shiny app.ShinyItemAnalysis
package was added.markdown
.DIF/Fairness
sections, histogram of total scores now better compares the
groups. THIS IS A CRAN VERSION
It includes versions 1.3.1-1 - 1.3.1-8 and following changes:
DIF/Fairness
section and in the Report
section.plotDIFLogistic()
now correctly plots matching criterion.DIF/Fairness
section now correctly plots best
fitting model when p.adjust.method is applied.DIF/Fairness
section now excludes missing values.LearningToLearn
was added to the package.LearningToLearn6
and LearningToLearn9
were added to the
app.plotDIFLogistic()
function was updated to accept class of Logistik
as input.plotDIFLogistic()
now accepts character input for item
argument.difNLR
, function ddfORD()
was renamed to difORD()
.DistractorAnalysis()
and plotDistractorAnalysis()
were unified.plotDIFLogistic()
it is now possible to turn off plotting empirical
probabilities. This can be done using draw.empirical = FALSE
.thr
. Default value is 0.2.plotDistractorAnalysis()
function when key
is missing.DIF/Fairness
section.plotAdjacent()
, plotCumulative()
and plotMultinomial()
were
fixed.DistractorAnalysis()
and plotDistractorAnalysis()
were updated to not requiring key
input.NA
values were fixed."key"
was unified.Science
were updated.plotMultinomial()
for data without NAs was fixed.DIF/Fairness
section.DIF/Fairness
section was updated.plotMultinomial()
was fixed. Function now displays more than 12
response patterns as linetype.plotAdjacent()
and plotCumulative()
were updated to
avoid removing data.THIS IS A CRAN VERSION
It includes versions 1.3.0-1 - 1.3.0-18 and following changes:
ItemAnalysis()
function, documentation for item reliability and
item validity indices has been added (Reported by Lot Fonteyne).plotDistractorAnalysis()
was fixed. X-ticks are now displayed
even with discrete = T argument.DistractorAnalysis()
and plotDistractorAnalysis()
now handle discrete
matching and groups specified via argument cut.points.DIF/Fairness
tab was fixed. Plots for items are now
displayed for dataBov dataset.difNLR
package version 1.2.3 plots for DDF are displayed
even for binary data. Plots were also fixed for data Bov.plotMultinomial()
was updated to return smooth curves
for any matching criterion.DIF/Fairness
section, histograms of Total scores for groups
are now properly displayed also for ordinal items.DistractorAnalysis()
was partially fixed, need to be revised.plotMultinomial()
was relaxed to decrease
size of reports.DistractorAnalysis()
was fixed. In case of possibly discrete
matching, matching is cutted based on its factors.plotCumulative()
which plots category probabilities
functions estimated by \code{multinom()} from nnet
package
was added. It is now used in application.plotAdjacent()
which plots category probabilities
functions estimated by \code{multinom()} from nnet
package
was added. It is now used in application.plotMultinomial()
which plots category probabilities
functions estimated by \code{multinom()} from nnet
package
was added. It is now used in application.DDplot()
now can displayed average item score as estimate of
difficulty for ordinal data. See argument average.score.DistractorAnalysis()
and plotDistractorAnalysis()
functions were
updated to allow for lowering the number of categories in cases when
cuts are not uniquely defined. Application was updated accordingly.Keep missing values
unfunctioning analyses were fixed
(missing data removed, warning message added)Keep missing values
warning message is now displayed
for those analyses for which missing values are omitted.DistractorAnalysis()
was fixed.startShinyItemAnalysis()
automatically launches default web browser.ggWrightMap()
now can display user-specified names of items
(Reported by Matthew G R Courtney)plotDIFLogistic()
now can display user-specified names
of reference and focal group.THIS IS A CRAN VERSION
It includes versions 1.2.9-1 - 1.2.9-3 and following changes:
ShinyItemAnalysis
in The R journal was added.plotDIFLogistic()
now handles user-specified matching
criterion.ItemAnalysis()
was added.Factor analysis
.Correlation structure
tab.Correlation structure
tab was updated.Reorganization of data upload and renaming data variables. test_answers()> nominal() correct_answers()> binary() scored_test()> total_score() test_key()> key() DIF_groups()> group() criterion_variable -> criterion()
And creating new variables: ordinal() ordinal data maximal() maximal values for ordinal data minimal() minimal values for ordinal data z_score() standardized total score, no need to use scale(total_score()) in each function
gDiscrim()
function has been extended to calculate Upper-Lower
Index (ULI) also for ordinal itemsplotDIFLogistic()
function was fixed. Function now returns
also item.name argument.plotDIFirt()
function was fixed. Function now returns also
item.name argument.THIS IS A CRAN VERSION
It includes versions 1.2.7-1 - 1.2.7-7
theme_app()
was created and applied for all
graphics, including R functions in package.THIS IS A CRAN VERSION
It includes versions 1.2.6-1 - 1.2.6-12
DDplot()
DIF/Fairness
section was updated.DIF/Fairness
section in report was improved.ShinyItemAnalysis
was created and added into reports.DIF/Fairness
IRT Lord and IRT Raju was extended to
1PL, 2PL and 3PL models.DIF/Fairness
IRT Lord and IRT Raju were fixed.plotDIFirt()
now can be also used for only one item.NLR IRT Z
DDF
plotDIFLogistic()
DDF
subsection of Fairness/DIF
section
was fixed.THIS IS A CRAN VERSION
difMedical
was renamed to MSAT-B
according to difNLR
package. Its description was mildly updated.Nonlinear IRT Z
and DDF
subsections
of DIF/Fairness
sections.DDF
analysis in Reports
section when customized
setting is selected.Reference
section and also in README file.Validity
section was mildly updated.ShinyItemAnalysis
now uses plotly
package!Report
page were updated.ShinyItemAnalysis
now used also data.table
package.gDiscrim()
to calculate generalized discrimination was
implemented.DDplot()
was edited to be able to handle generalized
discrimination.
Double slider was added into Traditional analysis page
.Summary
- Summary.RValidity
- Validity.RTraditional analysis
- TraditionalAnalysis.RRegression
- Regression.RIRT models
- IRT.RDIF/Fairness
- DIF.IRTDistractors
of Item analysis
page was mildly edited.Distractor
subpage. It is now
available on Traditional item analysis
subpage.ggplot2
package.THIS IS A CRAN VERSION
ShinyItemAnalysis
can now handle missing values. Newly there is
an option to handle NA values as zeros in scored data.dataMedical
, dataMedicalkey
, dataMedicaltest
were updated. Group variable and criterion variable were added.difNLR v 1.0.2
,
difR v 4.7
, ggplot2 v 2.2.1
, mirt v 1.24
, shiny 1.0.3
and
shinyjs v 0.9
).none
.GMAT
and GMATtest
from difNLR
package include now criterion
variable which can be used in Predictive validity section.DistractorAnalysis()
and plotDistractorAnalysis()
were
extended by matching
argument.difNLR
, mirt and ggplot2
.knitr
and xtable
are
now used in report generation).THIS IS A CRAN VERSION
dataMedical
, dataMedicalkey
, dataMedicaltest
were added into the package.difNLR
functions was edited to response to new version of
package.plotDIFirt()
was updated.THIS IS A CRAN VERSION
ggplot2
packagedifNLR
packageDDplot()
function for graphical representation
of difficulties and discriminations.DistractorAnalysis()
function and its graphical
representation plotDistractorAnalysis()
.plotDIFirt()
and plotDIFLogistic()
functions for
plotting characteristic curves based on IRT models
and based on logistic regression.Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.