++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
----------------------- CDM NEWS -----------------------------
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
CDM package
+-+-+-+
|C|D|M|
+-+-+-+
A. Robitzsch [aut, cre], T. Kiefer [aut],
A. C. George [aut], & A. Uenlue [aut]
mailto:
robitzsch@leibniz-ipn.de
For reporting a bug, please always provide a reproducible
R script and (if necessary) a corresponding dataset.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
--------------------------------------------------------------
VERSIONS CDM 7.5 | 2020-03-10 | Last: CDM 7.5-15
--------------------------------------------------------------
FIXED * corrected a bug that stopped gdina.dif()
(thanks to Sookhyun Park)
NOTE * included argument 'rule' in ideal.response.pattern().
Ideal response patterns can now be also computed for
the DINO model (in addition to the DINA model).
(suggested by Emine Yavuz)
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ideal.response.pattern(1)
--------------------------------------------------------------
VERSIONS CDM 7.4 | 2019-09-04 | Last: CDM 7.4-19
--------------------------------------------------------------
NOTE * included item content for dataset data.Student in the
?data.Students (requested by Ren Liu)
NOTE * corrected ?din (thanks @ifengji,
https://github.com/alexanderrobitzsch/CDM/pull/2)
NOTE * included NAs for non-defined entries for MD statistic
in IRT.RMSD()
NOTE * removed redundant code in gdina() function
(thanks to Jeremy Zheng; issue #4)
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
VERSIONS CDM 7.3 | 2019-03-18 | Last: CDM 7.3-17
--------------------------------------------------------------
NOTE * improved estimation with monotonicity constraints
(thanks to discussion with Xin Xu)
NOTE * changed normalization of penalty terms in regularized
estimation in gdina() and slca()
NOTE * fixed numerical instabilities for extreme probabilities
in reglca() (thanks to Boris Girnat)
NOTE * exported utility function cdm_attach_internal_function()
ADDED * added S3 method plot_item_mastery() for plotting item
response probabilities for non-masters and masters
DATA * included/modified datasets: ---
EXAMP * included/modified examples: plot_item_mastery (1)
--------------------------------------------------------------
VERSIONS CDM 7.2 | 2019-02-08 | Last: CDM 7.2-30
--------------------------------------------------------------
NOTE * changes in estimation algorithm in reglca()
NOTE * changes in estimation algorithm for regularized
estimation in gdina() and slca()
ADDED * included argument 'regular_weights' for adaptive
regularized estimation in gdina()
NOTE * added simulation function in data.dtmr
FIXED * fixed a bug in slca() in estimation problems with
equality constraints
NOTE * included data check routine for non-defined categories
in Q-matrix in mcdina() (suggested by Marcus Kubsch)
FIXED * fixed a bug in cdm.est.class.accuracy() when skill names
were missing in Q.matrix (reported by Sandip Sinharay)
FIXED * prevent negatively estimated probabilities in gdina()
function when PEM acceleration is used
(thanks to Yi-Jhen Wu)
ADDED * added S3 method IRT.marginal_posterior() for computing
marginal posterior distributions
(requested by Elif Oezlem Ardic)
ADDED * included checks of identifiability conditions of the
DINA model according Gu and Xu in din_identifiability()
DATA * included/modified datasets: data.dtmr, data.cdm09,
data.cdm10
EXAMP * included/modified examples: reglca (1), gdina (13),
data.dtmr (2), IRT.marginal_posterior (1),
din_identifiability (1)
--------------------------------------------------------------
VERSIONS CDM 7.1 | 2018-12-13 | Last: CDM 7.1-20
--------------------------------------------------------------
NOTE * included function numerical_gradient()
NOTE * changed computation of numerical gradient to symmetric
difference quotient in numerical_Hessian()
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
VERSIONS CDM 7.0 | 2018-11-06 | Last: CDM 7.0-12
--------------------------------------------------------------
NOTE * included argument 'prob_min' in reglca() to prevent
negatively estimated probabilities or numerical
instabilities (thanks to Boris Girnat)
NOTE * included utility function cat_paste()
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
VERSIONS CDM 6.6 | 2018-09-30 | Last: CDM 6.6-5
--------------------------------------------------------------
NOTE * changes in estimation algorithm in regularized latent
class analysis function reglca() to stabilize estimation
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
VERSIONS CDM 6.5 | 2018-08-16 | Last: CDM 6.5-12
--------------------------------------------------------------
NOTE * changed metric of regularization parameter lambda in
gdina() which seems to improve estimation stability
NOTE * exported utility function cdm_parameter_regularization(),
cdm_calc_information_criteria(), cdm_penalty_values(),
cdm_print_summary_information_criteria()
FIXED * corrected output with missing entries in case of groups
without observations in IRT.RMSD()
(thanks to Laura Helbling)
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
VERSIONS CDM 6.4 | 2018-07-06 | Last: CDM 6.4-23
--------------------------------------------------------------
NOTE * included argument 'optimizer' in gdina() which allows
to use the stats::optim() function for optimization in
M-step (default is optimizer='optim').
'use_optim=TRUE' is always used for estimation
of the RRUM model. This change was motivated by an
application of Hamdollah Ravand.
NOTE * included utility function cdm_matrixstring()
FIXED * allowed items with only zero Q-matrix entries in
mcdina() anf fixed a bug in reduced skill space
estimation in mcdina() (thanks to Jeannot Gersner)
NOTE * exported core computation function IRT_RMSD_calc_rmsd()
for RMSD statistic
DATA * included/modified datasets: ---
EXAMP * included/modified examples: gdina (13)
--------------------------------------------------------------
VERSIONS CDM 6.3 | 2018-05-28 | Last: CDM 6.3-45
--------------------------------------------------------------
NOTE * renamed thresholding penalty functions
NOTE * deleted seed argument from cdm.est.class.accuracy()
function
ADDED * included general simulation function sim_model()
ADDED * included function IRT.classify() for computing
individual classifications
ADDED * included function eval_likelihood() for evaluating
likelihood (and posterior) and prep_data_long_format()
for saving item response matrix in long format
NOTE * included corrected statistics for classification
accuracy and consistency in cdm.est.class.accuracy()
DATA * included/modified datasets: ---
EXAMP * included/modified examples: sim_model (1),
IRT.classify (1), eval_likelihood (1)
--------------------------------------------------------------
VERSIONS CDM 6.2 | 2018-05-02 | Last: CDM 6.2-91
--------------------------------------------------------------
FIXED * fixed a bug in IRT.RMSD() function for models fitted in
the CDM package
NOTE * speeded parts of the IRT.modelfit() function
ADDED * included item discrimination indices in discrim.index()
function (Henson, DiBello & Stout, 2018)
NOTE * included argument 'IDI_diff' in the din.validate.qmatrix()
function. Speeded this function due to referring most
of the computations to Rcpp.
ADDED * included cognitive diagnosis model for simultaneous
identification for skills and misconceptions (Kuo et al.,
2018, APM). The model is included in the gdina() function
using the argument rule='SISM'.
DATA * included/modified datasets: data.timss07.G4.py,
data.timss07.G4.Qdomains, data.timss11.G4.sa, data.cdm05,
data.cdm06, data.cdm07, data.pisa00R.ct, data.pisa00R.cc,
data.cdm08
EXAMP * included/modified examples: IRT.RMSD (3),
data.timss07.G4.lee (2), data.cdm (2), data.pisa00R (1),
gdina (12)
--------------------------------------------------------------
VERSIONS CDM 6.1 | 2018-02-26 | Last: CDM 6.1-10
--------------------------------------------------------------
FIXED * fixed bug of computation of MD statistic in IRT.RMSD()
which was recently broken
NOTE * included item index in title of figures in
IRT.irfprobPlot()
DATA * included/modified datasets: ---
EXAMP * included/modified examples: IRT.RMSD (2)
--------------------------------------------------------------
VERSIONS CDM 6.0 | 2017-11-20 | Last: CDM 6.0-101
--------------------------------------------------------------
ADDED * included regularized maximum likelihood in slca()
ADDED * included option for positivity constraints of Xlambda
parameter in slca()
ADDED * included P-EM acceleration method in slca(), gdm(),
and gdina()
NOTE * avoided inversion of singular covariance matrices in
gdina() by using MASS::ginv()
NOTE * included several sub-datasets and different Q-matrices
of the fraction subtraction dataset (see 'data.fraction')
NOTE * slight changes in algorithm of higher-order GDINA model
NOTE * included argument 'diag_only' in numerical_Hessian()
NOTE * exported utility functions cdm_fa1(), cdm_print_summary_call(),
cdm_print_summary_computation_time(), threshold operators
cdm_XXX_threshold() for diverse 'XXX', cdm_fit_normal(),
cdm_parameter_regularization(), cdm_penalty_values(),
CDM_rbind_fill(), cdm_matrix1(), cdm_matrix2()
NOTE * included derivative function numerical_Hessian_partial()
ADDED * added option for monotonicity constraints in gdina()
ADDED * added regularization methods in gdina(). Implemented are
the penalties SCAD, lasso, ridge, SCAD-L2, truncated L1, MCP
(Xu & Shang, 2017, JASA)
NOTE * included argument 'freq_weights' in gdina() to disable the
usage of frequency weights
ADDED * included prior distributions for item parameters in
gdina() function
ADDED * added regularized latent class estimation (reglca();
Chen et al., 2015, JASA; Chen et al., 2017, Psychometrika)
NOTE * extended option for skill space reduction in gdina() to
multiple group analysis
NOTE * enabled 'delta.designmatrix' argument in gdm() function
which was previously ignored
NOTE * disbaled PEM acceleration for HOGDINA models (not yet
implemented, thanks to Hamdollah Ravand)
DATA * included/modified datasets: data.fraction3,
data.fraction4, data.fraction5
EXAMP * included/modified examples: gdina (1.9, 1.10, 1.11),
slca (6), reglca (1)
--------------------------------------------------------------
VERSIONS CDM 5.9 | 2017-10-06 | Last: CDM 5.9-27
--------------------------------------------------------------
FIXED * fixed a bug in 'gdina' function in argument 'invariance'
if a vector of item names is used
(thanks to Hamdollah Ravand)
FIXED * fixed the broken function 'gdina.dif'
(thanks again to Hamdollah Ravand)
ADDED * added EM acceleration method P-EM in function 'slca' by
argument 'PEM'
NOTE * included P-EM utility functions 'cdm_pen_XXX' and
'cdm_print_summary_data_frame'
DATA * included/modified datasets: ---
EXAMP * included/modified examples: gdina (5.5)
--------------------------------------------------------------
VERSIONS CDM 5.8 | 2017-08-07 | Last: CDM 5.8-9
--------------------------------------------------------------
NOTE * fixed a bug in 'equivalent.dina' and included an
additional example (thanks to Sedat San)
NOTE * moved lattice and psych packages from 'Imports' to
'Suggests'. Removed MASS and plyr packages from
'Imports'.
NOTE * imported utility functions 'CDM_require_namespace'
and 'CDM_rmvnorm'.
FIXED * fixed a bug in 'gdina' for Q-matrices with only one
skill (thanks to Matthew Madison)
DATA * included/modified datasets: ---
EXAMP * included/modified examples: equivalent.dina (3)
--------------------------------------------------------------
VERSIONS CDM 5.7 | 2017-06-24 | Last: CDM 5.7-16
--------------------------------------------------------------
ADDED * included S3 method 'IRT.frequencies' for calculation
of observed and expected frequencies of univariate and
bivariate distributions (suggested by Rudolf Debelak)
DATA * included/modified datasets: ---
EXAMP * included/modified examples: IRT.frequencies (1)
--------------------------------------------------------------
VERSIONS CDM 5.6 | 2017-06-17 | Last: CDM 5.6-16
--------------------------------------------------------------
NOTE * internal restructure of 'gdina' and 'gdm' functions
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
VERSIONS CDM 5.5 | 2017-04-11 | Last: CDM 5.5-21
--------------------------------------------------------------
NOTE * corrected a wrong specification in the Rd manual
of 'skillspace.hierarchy' (Example 2; thanks to
Peida Zhan)
FIXED * corrected calculations of MD and MAD statistics in
'IRT.RMSD'
FIXED * fixed a bug in computation of 'itemfit.sx2' in case of
cells with frequencies of zero (thanks to Carmen Koehler)
DATA * included/modified datasets: ---
EXAMP * included/modified examples: skillspace.hierarchy (2)
--------------------------------------------------------------
CDM 5.4 (2017-02-08) -- CRAN CDM 5.4-0
--------------------------------------------------------------
FIXED * fixed warnings in calculation of standard errors in
'gdina' and 'gdina.wald'
(thanks to Andres Alberto Burga Leon)
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
CDM 5.3 (2017-01-11) -- CRAN CDM 5.3-0
--------------------------------------------------------------
ADDED * included function CDM::IRT.RMSD which computed group-wise
item fit. The RMSD item fit statistic was denoted as
RMSEA in previous publications and is internally computed
by the CDM::itemfit.rmsea function in the CDM package.
ADDED * included chi square statistic and MAD statistic (mean
absolute deviation) of item fit in CDM::IRT.RMSD
NOTE * cleaned code in function 'WaldTest' a bit
(thanks to suggestions of Michael Chirico)
NOTE * added links to recently released GDINA package in R.
Included Example 1d for fitting a GDINA model with
GDINA::GDINA.
DATA * included/modified datasets: ---
EXAMP * included/modified examples: IRT.RMSD (1), gdina (1d)
--------------------------------------------------------------
CDM 5.2 (2016-10-14) -- CRAN CDM 5.2-1
--------------------------------------------------------------
NOTE * extended 'item_by_group' function which can now handle
partial set of group-specific items
ADDED * included option for a subset of invariant items in
multiple group estimation of GDINA model in 'gdina'
(requested by Hamid Bahrami)
FIXED * fixed a bug in setting boundaries for guessing and slipping
parameters in 'din' (thanks to Jill-Jen Vie)
NOTE * reference to new Journal of Statistical Software George et al.
article in the CITATION file and the CDM manual
DATA * included/modified datasets: ---
EXAMP * included/modified examples: item_by_group (1),
gdina (5)
--------------------------------------------------------------
CDM 5.1 (2016-08-29) -- CRAN CDM 5.1-0
--------------------------------------------------------------
FIXED * changed the method of standard error calculation in
'gdina' function. It seems that previous CDM version produced
overestimated standard errors (thanks to Xue-Lan Qiu).
The improved calculation is now the default specified as the
argument 'se_version=1'. The behavior of older CDM versions can
be obtained with 'se_version=0'.
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
CDM 5.0 (2016-07-11) -- CRAN CDM 5.0-0
--------------------------------------------------------------
NOTE * changed the way of calculating numerical derivatives in
'numerical_Hessian'
NOTE * included utility function 'abs_approx' for a differentiable
function approximation (and its derivative 'abs_approx_D1')
of absolute value function
NOTE * added argument 'append' in function 'osink'
NOTE * included a utility function 'summary_sink' which sinks
summary output of a method into a file
NOTE * included utility function 'deltaMethod' for nonlinear
transformations of parameters
NOTE * included utility function 'item_by_group' which extends
a dataset of item responses and a grouping variable to an
extended dataset of group-specific item responses
ADDED * added sequential GDINA model for polytomous item responses
(Ma & de la Torre, 2016, BJMSP) in 'gdina' function.
NOTE * included argument 'invariance' in 'gdina' which allows
specification of non-invariant item parameters across
multiple groups (see ?gdina, Example 5)
NOTE * changed number of M-steps in 'slca' in default to 10
iterations (msteps=10)
DATA * included/modified datasets: data.cdm04
EXAMP * included/modified examples: abs_approx (1),
summary_sink (1), deltaMethod (1), item_by_group (1),
gdina (5,11)
--------------------------------------------------------------
CDM 4.991 (2016-05-31) -- CRAN CDM 4.991-1
--------------------------------------------------------------
NOTE * changed argument structure of 'numerical_hessian'
NOTE * included maximum likelihood estimation method for
GDINA models (method='ML') which seems to be more stable
when method='ULS' (or 'WLS') has a bad convergence
behavior. Now, the GDINA function switches to ML
estimation if the RRUM model is used
(thanks to a correspondence with Thomas Eckes).
NOTE * added an example of fitting (data.cdm, Example 1) the
reduced RUM model in the CDM package (see the recent
Chiu et al., IJT paper)
NOTE * changed default value of 'max.increment' argument in
'gdina' function to .3
FIXED * fixed a bug in 'gdina.dif' (thanks to Sinem Senfarah)
DATA * included/modified datasets: data.cdm03
EXAMP * included/modified examples: data.cdm (1)
--------------------------------------------------------------
CDM 4.8 (2016-03-10) -- CRAN CDM 4.8-0
--------------------------------------------------------------
FIXED * fixed a bug in 'itemfit.sx2' for fitted objects in TAM
NOTE * slightly changed output of 'print' methods
NOTE * changed argument 'rdigits' to 'digits' in 'summary'
method if necessary
NOTE * included utility functions 'osink' and 'csink' for
writing output of 'summary' into a file
NOTE * allowed argument 'file' in 'summary' method for objects
of class 'gdina', 'gdm', 'mcdina' and 'slca'
ADDED * included function 'numerical_Hessian' for a numerical
evaluation of the Hessian matrix
DATA * included/modified datasets: ---
EXAMP * included/modified examples: osink (1),
numerical_Hessian (1)
--------------------------------------------------------------
CDM 4.7 (2016-01-30) -- CRAN CDM 4.7-0
--------------------------------------------------------------
NOTE * extended Wald test for item fit 'gdina.wald' for DINO
models. In addition, an effect size of approximation
('uwgtdist' and 'wgtdist') is included
NOTE * cleaned namespace file and defined imported functions
in a more explicit way
ADDED * added a S3 method 'IRT.parameterTable'
ADDED * added a utility function 'WaldTest' for performing
Wald tests
DATA * included/modified datasets: ---
EXAMP * included/modified examples: gdina.wald (1), data.ecpe (1),
--------------------------------------------------------------
CDM 4.6 (2015-11-04) -- CRAN CDM 4.6-0
--------------------------------------------------------------
NOTE * included optional lower and upper bounds for estimation
of guessing and slipping parameters in 'din' (arguments
'guess.min', 'guess.max', 'slip.min', 'slip.max')
(requested by Kevin Carl Santos)
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
CDM 4.5 (2015-08-28) -- CRAN CDM 4.5-0
--------------------------------------------------------------
NOTE * extended 'din' and 'gdina' functions for items with more
than five measured attributes per item
(requested by Alvaro Artavia Medrano)
NOTE * included 'converged' value for all models indicating
whether maximum number of iterations was reached
(suggested by an anonymous reviewer)
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
CDM 4.4 (2015-07-01) -- CRAN CDM 4.4-1
--------------------------------------------------------------
FIXED * fixed a bug in function 'sequential.items'
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
CDM 4.3 (2015-06-05) -- CRAN CDM 4.3-0
--------------------------------------------------------------
NOTE * included t and p values in 'IRT.jackknife.gdina'
FIXED * fixed bugs in 'summary.gdina' and 'gdina' due to
computational problems in standard error calculation
DATA * included/modified datasets: ---
EXAMP * included/modified examples: ---
--------------------------------------------------------------
CDM 4.2 (2015-02-23) -- CRAN CDM 4.2-12
--------------------------------------------------------------
FIXED * fixed the ignorance of the 'centered.latent' argument
in the 'gdm' function
ADDED * included S3 methods 'IRT.expectedCounts' for extracting
expected counts and 'IRT.data' for extracting the used
item response dataset
ADDED * included general plotting function for item response
functions 'IRT.irfprobPlot'. This function can be used
with both lattice and graphics packages.
ADDED * included S3 method 'predict' (and function 'IRT.predict')
for computing expected values and residuals for
fitted models in CDM
NOTE * added attributes group and weights to 'IRT.data' value
NOTE * added a summary method for 'IRT.compareModels'
NOTE * changed output of posterior in 'gdina'
ADDED * included function 'IRT.repDesign' for constructing
replicate designs
ADDED * added arguments for fixing and initial parameters in
'gdina' function
NOTE * added parameter table and matrix of polychoric correlations
among skills in 'gdina' output
ADDED * added S3 method 'IRT.jackknife' and an initial version
for models of class 'gdina'
ADDED * added function 'IRT.derivedParameter' for replication methods
for derived parameters based on objects of class 'IRT.jackknife'
DATA * included/modified datasets: data.timss11.G4.AUT,
data.timss11.G4.AUT.part
EXAMP * included/modified examples: IRT.expectedCounts (1,2),
IRT.irfprobPlot (1,2), predict (1,2),
IRT.data (1), IRT.repDesign (1,2), gdina (1),
IRT.jackknife (1,2)
--------------------------------------------------------------
CDM 4.1 (2014-12-17) -- CRAN CDM 4.1
--------------------------------------------------------------
NOTE * included confidence intervals S3 method for din objects
FIXED * fixed a bug in printing BIC for 'gdina' objects
NOTE * included S3 method 'IRT.modelfit' for assessment of
absolute model fit
NOTE * included S3 method 'IRT.se' for computation of standard
errors
NOTE * included S3 method 'IRT.IC' for extracting information
criteria and 'IRT.compareModels' for model comparisons
NOTE * included object names in 'anova' methods for CDM objects
NOTE * included S3 method 'IRT.anova'
NOTE * removed MX2 statistic from 'modelfit.cor' function
DATA * included/modified datasets: ---
EXAMP * included/modified examples: vcov.din (1), IRT.modelfit (1),
IRT.IC (1), IRT.compareModels (1), anova (1,2)
--------------------------------------------------------------
CDM 4.0 (2014-11-21) -- CRAN CDM 4.0
--------------------------------------------------------------
NOTE * included argument for iteration progress in 'gdm'
and 'slca'
NOTE * included value call in 'din', 'gdina', 'gdm',
'mcdina' and 'slca'
NOTE * changed print method in 'din' and included print method
for 'gdina', 'gdm' and 'slca'
NOTE * added anova and coef method for 'slca' function
ADDED * added asymptotic covariance matrix estimation S3 method
'vcov' for objects of class din
ADDED * added computation of entropy of whole test and items in
function 'entropy.lca' (Asparouhov & Muthen, 2014)
NOTE * included a summary function for 'cdi.kli'
NOTE * included a function 'skillspace.full' for creating a
skill space for dichotomous skills
ADDED * included a plot S3 method for 'gdm' and 'slca'
NOTE * corrected a bug in person value of 'gdm': missing standard
errors of EAPs of multiple dimensions are now included
ADDED * included person fit appropriateness statistic (Liu et al., 2009)
ADDED * included S3 methods 'IRT.likelihood', 'IRT.personfit',
'IRT.irfprob' and 'IRT.factor.scores' for objects of classes
din, gdina, gdm, slca and mcdina
NOTE * included person parameter estimation (classification) for
'slca' and 'mcdina' functions
DATA * included/modified datasets: ---
EXAMP * included/modified examples: data.dcm (1), modelfit.cor (2),
vcov-CDM (1), entropy.lca (1,2,3), personfit.appropriateness (1,2),
slca (1), IRT.likelihood (1), IRT.irfprob (1), IRT.factor.scores (1)
--------------------------------------------------------------
CDM 3.4 (2014-08-31) -- CRAN CDM 3.4-21
--------------------------------------------------------------
NOTE * speeded itemfit.sx2 a bit
NOTE * included output value 'pattern' in multiple group GDINA models
(requested by Sung Eun Kim)
NOTE * included adjusted mean absolute deviation of Q3 statistics (MADaQ3)
in modelfit.cor.din
EXAMP * included examples for running CDMs in mirt package
o DINA model ('data.dcm', Model 1)
o additive GDINA model ('data.dcm', Model 6)
DATA * included/modified datasets: ---
EXAMP * included/modified examples: modelfit.cor (1), din.deterministic,
itemfit.sx2 (1), data.dcm (1,6)
--------------------------------------------------------------
CDM 3.3 (2014-05-25) -- CRAN CDM 3.3-24
--------------------------------------------------------------
ADDED * included structured latent class analysis for polytomous data ('slca')
DATA * included/modified datasets: ---
EXAMP * included/modified examples: slca (1,2,3,4,5)
--------------------------------------------------------------
CDM 3.2 (2014-04-23) -- CRAN CDM 3.2-6
--------------------------------------------------------------
NOTE * included an argument save.devmin in 'gdina' for more parsimonious
savings of output values to avoid working memory overflow for
large datasets
NOTE * included effect size of differential item functioning in
gdina model (included in 'gdina.dif')
FIXED * fixed a bug in itemfit.sx2
DATA * included/modified datasets: ---
EXAMP * included/modified examples: gdina.dif (1)
--------------------------------------------------------------
CDM 3.1 (2014-03-21) -- CRAN CDM 3.1-14
--------------------------------------------------------------
FIXED * fixed a bug when printing objects of class 'din'
NOTE * allowed vector input of item response pattern in 'plot.din'
NOTE * disabled argument log.file in 'summary.din'
NOTE * included random initial parameter values in 'din' and 'gdina'
NOTE * changed some settings in plot.din
DATA * included/modified datasets: ---
EXAMP * included/modified examples: din (1), gdina (1), plot.din
--------------------------------------------------------------
CDM 3.0 (2014-02-26) -- CRAN CDM 3.0-29
--------------------------------------------------------------
NOTE * changed output in summary.gdina a bit (fixed a bug for
the output of skill probabilities)
NOTE * included the argument 'avoid.zeroprobs' in gdina function
to avoid zero probabilities of skill classes and inversion
of singular matrices
NOTE * fixed problems in itemfit.sx2 for empty cells (thanks to
Balal Izanloo)
ADDED * included a Wald test for the GDINA model (gdina.wald)
ADDED * included assessment of item-wise differential item functioning
in the GDINA model (gdina.dif)
FIXED * removed minor bug in summary.gdina
FIXED * fixed a bug with singular design matrix in skill space reduction
ADDED * included calculation of polychoric correlation in summary.gdina
and an additional function skill.polychor
NOTE * removed parallel plot form plot.din
FIXED * fixed a crucial bug for in the multidimensional GDMs with item slope
parameter estimation
ADDED * included function 'mcdina' for estimating the multiple group multiple
choice DINA (MCDINA) model with polytomous attributes
DATA * included/modified datasets: data.cdm01, data.cdm02
EXAMP * included/modified examples: data.ecpe, gdina.wald (1),
gdina.dif (1), din (7), mcdina (1,2,3)
--------------------------------------------------------------
CDM 2.7 (2014-01-30) -- CRAN CDM 2.7-7
--------------------------------------------------------------
ADDED * included cognitive diagnostic indices based on Kullback-Leibler
information ('cdi.kli')
NOTE * changed NAMESPACE and DESCRIPTION files due to a change in Rcpp
NOTE * included non-exported Rcpp files for calculating individual
likelihood and posterior distribution for polytomous items and
multiple groups
DATA * included/modified datasets: ---
EXAMP * included/modified examples: cdi.kli (1)
--------------------------------------------------------------
CDM 2.6 (2014-01-19) -- CRAN CDM 2.6-13
--------------------------------------------------------------
FIXED * fixed a bug in polytomous GDINA models with only one skill
NOTE * included value 'zeroprob.skillclasses' in skillspace.hierarchy
NOTE * included a more efficient skill space reduction method in
gdina: using the argument 'reduced.skillspace.method',
the new implemented method is now the default
ADDED * included data transformation function 'sequential.items' for
constructing dichtomous pseudo items from polytomous items for
estimation sequential item response models
DATA * included/modified datasets: data.sda6, data.timss03.G8.su,
data.timss07.G4.lee
EXAMP * included/modified examples: gdm (10)
--------------------------------------------------------------
CDM 2.5 (2014-01-16) -- CRAN CDM 2.5-48
--------------------------------------------------------------
FIXED * fixed a bug when inverting a singular matrix in standard error
calculation in gdina function (thanks to Hsin-Hao Chen)
NOTE * included conditional item probabilities in gdina output
NOTE * included gdina plot function (thanks to a discussion with
Hamdollah Ravand)
FIXED * fixed a bug in summary.dina if data without column names
is provided (thanks to Hu Jinxiang)
FIXED * removed a bug in gdina function which occured sometimes
due to division by zero (thanks to Hu Jinxiang)
FIXED * fixed a bug in cdm.est.class.accuracy for GDINA models
(thanks to Hamdollah Ravand)
FIXED * fixed some instable numerical behavior for higher order
GDINA models (thanks to Ed Merkle)
FIXED * fixed a bug in calculating number of estimated parameters for
'din' function in case of skill classes with zero probability
(thanks to Juergen Gross)
NOTE * added some explanations in the manual of 'modelfit.cor'
ADDED * included S-X2 item fit statistic ('itemfit.sx2') accompanied
with a summary and a plot method
ADDED * included marginal skill classification reliability in
'cdm.est.class.accuracy' function
ADDED * implemented significance tests in modelfit.cor and modelfit.cor.din
ADDED * included jackknife estimation of standard
errors and bias corrected estimators in modelfit.cor.din
NOTE * speeded function 'modelfit.cor' by moving some parts of the code
to Rcpp
NOTE * included warnings in case of nonconvergence in gdina function
(thanks to a dataset provided by Hamdollah Ravand)
NOTE * slight change of algorithm for skill space reduction which
seems to improve convergence behavior
ADDED * included a function for skill space approximation
ADDED * included a function for calculation reduced skill spaces
if hierarchies among skills exist ('skillspace.hierarchy')
ADDED * included generalized distance discriminating method and a function
for computing ideal response patterns
DATA * included/modified datasets: data.hr, data.melab, data.jang,
data.dtmr
EXAMP * included/modified examples: gdm (4,5,6), gdina (8,9), sim.din (3),
sim.gdina (2,3), itemfit.sx2 (1,2), din(6)
--------------------------------------------------------------
CDM 2.4 (2013-10-24) -- CRAN CDM 2.4-9
--------------------------------------------------------------
DATA * included two datasets 'data.dcm' and 'data.ecpe'
DATA * eliminated a recoding error in data.Students
--------------------------------------------------------------
CDM 2.3 (2013-09-06) -- CRAN CDM 2.3-15
--------------------------------------------------------------
NOTE * worked on slight improvements of computation time in din
and gdina
DATA * included multiple group dataset 'data.mg'
NOTE * cleaned CDM manual (thanks to Michaela Mayr)
--------------------------------------------------------------
CDM 2.2 (2013-08-20) -- CRAN CDM 2.2-20
--------------------------------------------------------------
NOTE * accelerated rowMaxs function
ADDED * included Q-matrix validation function (de la Torre, 2008)
ADDED * included joint maximum likelihood estimation and
classification function for the mixed DINA/DINO model
-> function din.deterministic
EXAMP * included another example (based on attribute vectors)
in sim.din function
ADDED * included polytomous GDINA model (Chen & de la Torre, 2013)
--------------------------------------------------------------
CDM 2.01 (2013-07-27) -- CRAN CDM 2.01-11
--------------------------------------------------------------
FIXED * fixed a bug in gdm in multiple group estimation: group
labeling was incorrect and is now included in output group.stat
and displayed in summary.gdm
NOTE * improved convergence behavior for multiple group analysis
NOTE * accelerated calculation in gdm: skillspace reduction
--------------------------------------------------------------
CDM 1.5 (2013-07-09) -- CRAN CDM 1.5-12
--------------------------------------------------------------
NOTE * allowed for centering of intercepts and slopes in gdm
ADDED * included satured skill distribution estimation in gdm
FIXED * fixed a bug in the calculation of RMSEA itemfit, although
is not clear if the correct formula is used
FIXED * fixed a bug in gdm with varying number of item categories
(thanks to Artur Pokropek)
ADDED * included the multidimensional latent class IRT model
(Bartolucci, 2007) in gdm (skillspace="est")
NOTE * speeded calculation of modelfit.cor
--------------------------------------------------------------
CDM 1.4 (2013-06-30) -- CRAN CDM 1.4-16
--------------------------------------------------------------
FIXED * eliminated a bug in GDINA function for link="log"
ADDED * included the parametrization of the reduced RUM model in gdina
ADDED * worked on higher order GDINA model which has the argument
HOGDINA in gdina. Now there is dependency on the 'fa' factor
analysis function in the psych package.
ADDED * included log-likelihood extractor functions
ADDED * included group-wise calculation of RMSEA statistics
see itemfit.rmsea and the gdm output (thanks to Artur Pokropek)
FIXED * (hopefully) fixed a bug in gdm function with differing number of
item categories (thanks to Artur Pokropek)
NOTE * use frequency patterns in gdm to speed up calculations
for large data sets (argument 'use.freqpatt' in gdm function)
NOTE * included more than one step in gdm function
- arguments msteps (set to 8 M steps as the default) and convM
--------------------------------------------------------------
CDM 1.3 (2013-05-15) -- CRAN CDM 1.3-17
--------------------------------------------------------------
ADDED * included the general diagnostic model (function gdm)
ADDED * included a function for calculating equivalent skill classes
for the DINA/DINO model
ADDED * modelfit and anova functions are also accessible for the GDM
FIXED * fixed a bug in the function for classification accuracy
ADDED * included the RMSEA item fit index
FIXED * fixed a bug in gdina estimation
ADDED * included coef extractor function
--------------------------------------------------------------
CDM 1.2-3 (2013-02-12)
--------------------------------------------------------------
ADDED * included a model fit function modelfit.cor(.din) for the
comparison of observed and expected correlations
FIXED * fixed a bug in the gdina function when applying the model
with one skill
--------------------------------------------------------------
CDM 1.2-1/1.2-2 (2013-02-01)
--------------------------------------------------------------
ADDED * allow for user specified reduced skill space designs
ADDED * included tetrachoric correlations in summary.gdina function
ADDED * handled zero probability classes in gdina function
ADDED * allowed for output and input of attribute patterns in sim.din
function
NOTE * included relative deviance change as convergence criterion in din
function
ADDED * included function for determining a statistically equivalent
DINA model (von Davier, 2013)
NOTE * changed structure of din summary
ADDED * included an anova function
ADDED * included a function for estimating classification reliability
FIXED * fixed a bug in gdina function when creating design matrices
--------------------------------------------------------------
CDM 1.2.0 (2012-09-30)
--------------------------------------------------------------
ADDED * implemented the latent skill space reduction according
to Xu and von Davier (2008)
FIXED * fixed a bug in calculation of estimated parameters
FIXED * fixed a bug in IDI calculation
FIXED * corrected MLE based classification in case of latent
classes with a priori spceified zero probability in the
din function (yet to be done in gdina)
FIXED * fixed a bug in standard error calculation in gdina
NOTE * speed improvements in din and gdina function
- included a C function for calculating the posterior
distribution
NOTE * check of admissible rule in gdina
ADDED * include rule "GDINA2" and "GDINA1" ("GDINA1" is
equivalent to "ACDM")
ADDED * Simulation function for the GDINA model included
ADDED * multiple group gdina function for estimation of
generalized DINA models (de la Torre, 2011)
ADDED * included new function summary.gdina
--------------------------------------------------------------
CDM 1.1 (2012-07-29)
--------------------------------------------------------------
NOTE * highest argument deleted from print.din method
NOTE * new argument for summary.din n.skill.classes
NOTE * skill pattern probabilities deleted from print.din
NOTE * skill pattern summary replaced by n most frequent skill
classes in summary.din
NOTE * (argument ask=TRUE in summary.din placed by overwrite=FALSE)
NOTE * use a different item discrimination index (Lee et al., 2012)
NOTE * remove GDINA estimation from this package version because
it is in an experimental status
-> gdina will be included in CDM >= 1.1-1
NOTE * new optional attribute for q-matrix specification list of skill
labels
NOTE * changed ouput in din function
-> pattern, like and posterior are sorted according to the original
input data frame
FIXED * remove a bug for models with only attribute
ADDED * new argument for din method zeroprob.skillclasses
ADDED * new argument for din method equal.itempars
NOTE * new arguments for summary.din method log.file and ask
FIXED * minor bug fixes in print.din method
NOTE * replace internal method rowProds with faster rowProds2
NOTE * option for writing an output file of model results included in
function summary.din
--------------------------------------------------------------
CDM 1.0 (2011-10-27)
--------------------------------------------------------------
[--- FIRST CRAN RELEASE ---]