pith. sign in

arxiv: 2605.05684 · v1 · submitted 2026-05-07 · 📊 stat.ME

Latent Impact and Differential Item Functioning Analysis for Asymmetric IRT Models

Gabriel Wallin , Qi Huang This is my paper

Pith reviewed 2026-05-08 08:06 UTC · model grok-4.3

classification 📊 stat.ME
keywords differential item functioninglatent classesasymmetric item response theorymixture modelsl1 regularizationmeasurement invarianceimpact analysiseducational testing
0
0 comments X

The pith

An l1-regularized mixture model identifies latent impact and DIF items in asymmetric IRT models without observed groups or anchors.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper introduces a framework to jointly model latent population differences and differential item functioning in item response theory when groups are unobserved and anchors unknown. It represents group differences via latent classes in a mixture model and uses an l1 penalty on item-specific DIF parameters to select which items function differently, assuming few such items. The method works with asymmetric link functions to better fit real response data. Simulations show good recovery of parameters, and applications to educational tests demonstrate its ability to distinguish impact from DIF.

Core claim

By combining latent class mixture modeling with asymmetric IRT and l1 regularization for DIF selection, the estimator simultaneously recovers the latent classes representing impact and identifies the sparse set of DIF items, all without requiring observed group labels or pre-specified anchor items.

What carries the argument

The ℓ1-regularised estimator applied to the DIF shift parameters in a latent class asymmetric IRT mixture model, which induces sparsity to select DIF items while estimating class-specific impact.

If this is right

  • The method recovers latent class probabilities, item parameters, and DIF effects accurately in simulations across varied configurations.
  • Applications to educational data can distinguish cases with both impact and DIF from those with mainly impact and little DIF.
  • Analysis of measurement invariance becomes possible in settings lacking observed groups or known anchors.
  • The framework accommodates asymmetric response processes that symmetric IRT models cannot capture well.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • This could be extended to other latent variable models in psychology or sociology where unobserved heterogeneity affects item responses.
  • Sensitivity analyses for the sparsity level would be needed if the few-DIF-items assumption is uncertain in a new dataset.
  • Pairing the estimator with cross-validation on held-out responses could help guard against over-selection of DIF items.

Load-bearing premise

The number of items with differential functioning is relatively small.

What would settle it

A simulation where the true proportion of DIF items is large would show whether the estimator still recovers the latent classes and correctly selects the DIF items or instead overfits.

Figures

Figures reproduced from arXiv: 2605.05684 by Gabriel Wallin, Qi Huang.

Figure 1
Figure 1. Figure 1: Bias of the item parameters for N = 1000 across values of π view at source ↗
Figure 2
Figure 2. Figure 2: RMSE of the item parameters for N = 1000 across values of π. 0.08–0.09 when π = 0.10 to about 0.22–0.23 when π = 0.50, with little variation across sample sizes. Note though that the performance relative to the naive classifier that puts every respondent in the baseline group shows clear improvement as the proportion in the second class increases. The AUC values are stable for N = 1000 and N = 3000, remain… view at source ↗
Figure 3
Figure 3. Figure 3: Bias of the item parameters for N = 1000 under Design B across values of π. π = 0.1 π = 0.3 π = 0.5 d δ 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0.00 0.04 0.08 0.12 0.00 0.04 0.08 0.12 Item RMSE view at source ↗
Figure 4
Figure 4. Figure 4: RMSE of the item parameters for N = 1000 under Design B across values of π. 18 view at source ↗
Figure 5
Figure 5. Figure 5: ROC curve for classifying country membership using posterior class probabilities view at source ↗
Figure 6
Figure 6. Figure 6: Density of posterior probabilities of belonging to the focal class, by country. view at source ↗
Figure 7
Figure 7. Figure 7: Bias of the item parameters for N = 500 across values of π under Design A. π = 0.1 π = 0.3 π = 0.5 d δ 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0.0 0.2 0.4 0.6 0.0 0.2 0.4 0.6 Item RMSE view at source ↗
Figure 8
Figure 8. Figure 8: RMSE of the item parameters for N = 500 across values of π under Design A. 35 view at source ↗
Figure 9
Figure 9. Figure 9: Bias of the item parameters for N = 3000 across values of π under Design A. π = 0.1 π = 0.3 π = 0.5 d δ 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 0.04 0.08 0.12 0.04 0.08 0.12 Item RMSE view at source ↗
Figure 10
Figure 10. Figure 10: RMSE of the item parameters for N = 3000 across values of π under Design A. 36 view at source ↗
Figure 11
Figure 11. Figure 11: Bias of the item parameters for N = 500 under Design B across values of π. 37 view at source ↗
Figure 12
Figure 12. Figure 12: RMSE of the item parameters for N = 500 under Design B across values of π. π = 0.1 π = 0.3 π = 0.5 d δ 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 −0.02 −0.01 0.00 0.01 0.02 −0.02 −0.01 0.00 0.01 0.02 Item Bias view at source ↗
Figure 13
Figure 13. Figure 13: Bias of the item parameters for N = 3000 under Design B across values of π. 38 view at source ↗
Figure 14
Figure 14. Figure 14: RMSE of the item parameters for N = 3000 under Design B across values of π. 39 view at source ↗
read the original abstract

Differential item functioning (DIF) arises alongside latent population heterogeneity in many applications, and both must be accounted for when assessing measurement invariance. In many practical settings, however, the comparison groups are unobserved and anchor items are unknown. A further challenge is that item response theory models traditionally assume symmetric link functions, yet empirical response processes may exhibit substantial asymmetry. This paper proposes a general framework for jointly analysing impact and DIF under asymmetric item response models. Unobserved group differences are represented by latent classes within a mixture item response model, while item-specific shifts capture DIF effects. Assuming the number of DIF items is relatively small, an $\ell_1$-regularised estimator is used to simultaneously identify the latent classes and select DIF items without requiring observed group labels or pre-specified anchor items. A simulation study evaluates recovery of impact, item parameters, and DIF effects across a range of configurations. The method is illustrated using two empirical applications from educational testing. In one dataset, the selected model reveals both impact and item-level DIF, whereas in the other, the results indicate substantial impact but little evidence of item-level DIF.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper proposes a mixture IRT framework for asymmetric link functions that jointly models latent impact (unobserved group heterogeneity) and DIF via item-specific shifts. An ℓ1-regularized estimator simultaneously identifies the latent classes and selects DIF items without observed group labels or pre-specified anchors, under the assumption that the number of DIF items is relatively small. Recovery is assessed in a simulation study across configurations, and the method is applied to two educational testing datasets, one showing both impact and DIF and the other showing impact with little DIF.

Significance. If the recovery and selection properties hold, the work offers a practical advance for measurement invariance analysis in settings with latent heterogeneity and asymmetric responses, where traditional methods requiring observed groups or anchors are infeasible. The joint estimation and regularization approach could improve fairness assessments in testing applications.

major comments (3)
  1. [§4] §4 (Simulation Study): recovery claims for impact, item parameters, and DIF effects are presented without error bars, standard errors, or detailed metrics on false positive/negative rates for DIF selection; this weakens assessment of the estimator's reliability across the tested configurations.
  2. [§3] §3 (Estimation procedure): the ℓ1-regularized estimator and class identification rest on the assumption that the number of DIF items is relatively small (stated in the abstract and method); no sensitivity analysis or robustness checks to violations of this assumption are reported, which is load-bearing for the central claim of simultaneous identification without anchors.
  3. [§3] §3, Eq. for the penalized likelihood: the regularization parameter and number of latent classes are treated as tuning parameters, but the manuscript provides limited guidance on their selection procedure and its effect on the resulting class recovery and DIF selection.
minor comments (2)
  1. [Abstract] Abstract: the two empirical datasets are described only generically as 'educational testing'; adding brief identifiers or sample sizes would improve context.
  2. [Notation] Notation: ensure consistent use of symbols for the asymmetric link function and the mixture weights across sections and equations.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive comments, which help clarify important aspects of the simulation study and estimation procedure. We respond to each major comment below and indicate the planned revisions.

read point-by-point responses

  1. Referee: [§4] §4 (Simulation Study): recovery claims for impact, item parameters, and DIF effects are presented without error bars, standard errors, or detailed metrics on false positive/negative rates for DIF selection; this weakens assessment of the estimator's reliability across the tested configurations.

    Authors: We agree that the simulation results would be strengthened by including variability measures and explicit selection performance metrics. In the revised manuscript, we will add standard deviations (or error bars) across the Monte Carlo replications for all reported recovery metrics on impact, item parameters, and DIF effects. We will also report false positive and false negative rates for DIF item selection under each configuration to provide a fuller picture of reliability. revision: yes

  2. Referee: [§3] §3 (Estimation procedure): the ℓ1-regularized estimator and class identification rest on the assumption that the number of DIF items is relatively small (stated in the abstract and method); no sensitivity analysis or robustness checks to violations of this assumption are reported, which is load-bearing for the central claim of simultaneous identification without anchors.

    Authors: The sparsity assumption on the number of DIF items is indeed central to the ℓ1-regularization strategy and the claim of anchor-free identification. We will add a sensitivity analysis in the simulation study (or an appendix) that varies the proportion of DIF items to examine how performance degrades when the assumption is moderately violated, thereby clarifying the practical scope of the method. revision: yes

  3. Referee: [§3] §3, Eq. for the penalized likelihood: the regularization parameter and number of latent classes are treated as tuning parameters, but the manuscript provides limited guidance on their selection procedure and its effect on the resulting class recovery and DIF selection.

    Authors: We will expand the relevant section to give more explicit guidance on selecting the regularization parameter and the number of latent classes. The revision will describe the concrete procedure employed (e.g., information criteria or cross-validation), report the chosen values, and include a brief discussion or supplementary results illustrating the sensitivity of class recovery and DIF selection to these choices. revision: yes

Circularity Check

0 steps flagged

No significant circularity; estimator builds on standard mixture and regularization methods

full rationale

The paper proposes a joint framework for latent impact and DIF under asymmetric IRT models using a mixture model with latent classes and an ℓ1-regularized estimator for simultaneous class identification and DIF item selection. This relies on established statistical techniques (mixture IRT, lasso-type regularization) without any derivation step that reduces a claimed result to a fitted parameter or self-citation by construction. The central claim (recovery of impact, parameters, and DIF without observed labels or anchors) is evaluated via simulation and empirical application rather than being tautological with the inputs. No self-definitional, fitted-input-as-prediction, or load-bearing self-citation patterns appear in the provided abstract or description. The assumption that the number of DIF items is small is stated explicitly as a modeling choice, not smuggled in as a derived result.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

The central claim rests on the assumption that DIF is sparse (small number of items), which justifies the ℓ1 penalty for simultaneous class identification and item selection. Standard mixture IRT assumptions for latent classes and asymmetric link functions are also invoked.

free parameters (2)
  • ℓ1 regularization parameter
    Controls sparsity level for DIF selection; value is chosen or tuned but not specified in abstract.
  • number of latent classes
    Determines the structure of unobserved groups; assumed or selected as part of the model.
axioms (2)
  • ad hoc to paper The number of DIF items is relatively small
    Explicitly stated in abstract to enable the ℓ1-regularized estimator to identify and select DIF items.
  • domain assumption Latent classes can represent unobserved group differences in response processes
    Core modeling choice for handling impact without observed labels.

pith-pipeline@v0.9.0 · 5485 in / 1336 out tokens · 94717 ms · 2026-05-08T08:06:32.859132+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

151 extracted references · 151 canonical work pages

  1. [1]

    and Cohen, Allan S

    Bolt, Daniel M. and Cohen, Allan S. and Wollack, James A. , title =. Journal of Educational Measurement , year =

    work page

  2. [2]

    British Journal of Mathematical and Statistical Psychology , year=

    Defining asymmetry in item response theory , author=. British Journal of Mathematical and Statistical Psychology , year=

    work page

  3. [3]

    British Journal of Mathematical and Statistical Psychology , volume=

    Identifiability analysis of the fixed-effects one-parameter logistic positive exponent model , author=. British Journal of Mathematical and Statistical Psychology , volume=. 2025 , publisher=

    work page 2025

  4. [4]

    Large-scale Assessments in Education , volume=

    Investigating item complexity as a source of cross-national DIF in TIMSS math and science , author=. Large-scale Assessments in Education , volume=. 2024 , publisher=

    work page 2024

  5. [5]

    The Journal of Machine Learning Research , volume=

    Model selection through sparse maximum likelihood estimation for multivariate Gaussian or binary data , author=. The Journal of Machine Learning Research , volume=. 2008 , publisher=

    work page 2008

  6. [6]

    Structural Equation Modeling: a Multidisciplinary Journal , volume=

    Simplifying the assessment of measurement invariance over multiple background variables: Using regularized moderated nonlinear factor analysis to detect differential item functioning , author=. Structural Equation Modeling: a Multidisciplinary Journal , volume=. 2020 , publisher=

    work page 2020

  7. [7]

    Psychometrika , volume=

    A statistical test for differential item pair functioning , author=. Psychometrika , volume=. 2015 , publisher=

    work page 2015

  8. [8]

    , author=

    Improving the assessment of measurement invariance: Using regularization to select anchor items and identify differential item functioning. , author=. Psychological Methods , volume=. 2020 , publisher=

    work page 2020

  9. [9]

    Journal of Educational Measurement , volume=

    Detection of test collusion via Kullback--Leibler divergence , author=. Journal of Educational Measurement , volume=. 2013 , publisher=

    work page 2013

  10. [10]

    Applied Psychological Measurement , volume=

    Graph theory approach to detect examinees involved in test collusion , author=. Applied Psychological Measurement , volume=. 2021 , publisher=

    work page 2021

  11. [11]

    Journal of Educational and Behavioral Statistics , volume=

    Measuring student ability, classifying schools, and detecting item bias at school level, based on student-level dichotomous items , author=. Journal of Educational and Behavioral Statistics , volume=. 2014 , publisher=

    work page 2014

  12. [12]

    Statistical Theories of Mental Test Scores , pages=

    Some latent trait models and their use in inferring an examinee's ability , author=. Statistical Theories of Mental Test Scores , pages=. 1968 , address=

    work page 1968

  13. [13]

    Marginal maximum likelihood estimation of item parameters: Application of an

    Bock, R Darrell and Aitkin, Murray , journal=. Marginal maximum likelihood estimation of item parameters: Application of an. 1981 , publisher=

    work page 1981

  14. [14]

    Multiple group

    Bock, R Darrell and Zimowski, Michele F , booktitle=. Multiple group. 1997 , publisher=

    work page 1997

  15. [15]

    Journal of Educational and Behavioral Statistics , volume=

    A mixture item response model for multiple-choice data , author=. Journal of Educational and Behavioral Statistics , volume=. 2001 , publisher=

    work page 2001

  16. [16]

    Applied Psychological Measurement , volume=

    An iterative procedure for linking metrics and assessing item bias in item response theory , author=. Applied Psychological Measurement , volume=. 1988 , publisher=

    work page 1988

  17. [17]

    Educational and Psychological Measurement , volume=

    A Monte Carlo study of an iterative Wald test procedure for DIF analysis , author=. Educational and Psychological Measurement , volume=. 2017 , publisher=

    work page 2017

  18. [18]

    arXiv preprint arXiv:2108.08604 , year=

    Item Response Theory--A Statistical Framework for Educational and Psychological Measurement , author=. arXiv preprint arXiv:2108.08604 , year=

    work page arXiv

  19. [19]

    Bayesian Analysis , volume=

    A Skew Item Response Model , author=. Bayesian Analysis , volume=

    work page

  20. [20]

    Journal of Educational and Behavioral Statistics , volume=

    Bayesian estimation of the logistic positive exponent IRT model , author=. Journal of Educational and Behavioral Statistics , volume=. 2010 , publisher=

    work page 2010

  21. [21]

    Journal of Educational and Behavioral Statistics , pages=

    Item pool quality control in educational testing: change point model, compound risk, and sequential detection , author=. Journal of Educational and Behavioral Statistics , pages=. 2022 , publisher=

    work page 2022

  22. [22]

    psychometrika , volume=

    DIF statistical inference without knowing anchoring items , author=. psychometrika , volume=. 2023 , publisher=

    work page 2023

  23. [23]

    Psychometrika , volume=

    Joint maximum likelihood estimation for high-dimensional exploratory item factor analysis , author=. Psychometrika , volume=. 2019 , publisher=

    work page 2019

  24. [24]

    The Annals of Applied Statistics , volume=

    Detection of two-way outliers in multivariate data and application to cheating detection in educational tests , author=. The Annals of Applied Statistics , volume=. 2022 , publisher=

    work page 2022

  25. [25]

    Cho, Sun-Joo and Suh, Youngsuk and Lee, Woo-yeol , journal=. An. 2016 , publisher=

    work page 2016

  26. [26]

    A multilevel mixture

    Cho, Sun-Joo and Cohen, Allan S , journal=. A multilevel mixture. 2010 , publisher=

    work page 2010

  27. [27]

    2017 , publisher=

    Handbook of quantitative methods for detecting cheating on tests , author=. 2017 , publisher=

    work page 2017

  28. [28]

    The effects of purification of matching criterion on the identification of

    Clauser, Brian and Mazor, Kathy and Hambleton, Ronald K , journal=. The effects of purification of matching criterion on the identification of. 1993 , publisher=

    work page 1993

  29. [29]

    Journal of Educational Measurement , volume=

    A mixture model analysis of differential item functioning , author=. Journal of Educational Measurement , volume=. 2005 , publisher=

    work page 2005

  30. [30]

    Detecting intervention effects using a multilevel latent transition analysis with a mixture

    Cho, Sun-Joo and Cohen, Allan S and Bottge, Brian , journal=. Detecting intervention effects using a multilevel latent transition analysis with a mixture. 2013 , publisher=

    work page 2013

  31. [31]

    Applied Psychological Measurement , volume=

    Latent transition analysis with a mixture item response theory measurement model , author=. Applied Psychological Measurement , volume=. 2010 , publisher=

    work page 2010

  32. [32]

    After differential item functioning is detected:

    Cho, Sun-Joo and Suh, Youngsuk and Lee, Woo-yeol , journal=. After differential item functioning is detected:. 2016 , publisher=

    work page 2016

  33. [33]

    International Journal of Pure and Applied Mathematics , volume=

    On Mixture Regression Shrinkage and Selection Via the MR-Lasso , author=. International Journal of Pure and Applied Mathematics , volume=

    work page

  34. [34]

    Journal of the American Statistical Association , volume=

    The adaptive lasso and its oracle properties , author=. Journal of the American Statistical Association , volume=. 2006 , publisher=

    work page 2006

  35. [35]

    Statistical Science , volume=

    In defense of the indefensible: A very naive approach to high-dimensional inference , author=. Statistical Science , volume=. 2021 , publisher=

    work page 2021

  36. [36]

    The Annals of Statistics , volume=

    Nearly unbiased variable selection under minimax concave penalty , author=. The Annals of Statistics , volume=

    work page

  37. [37]

    Journal of the American statistical Association , volume=

    Variable selection via nonconcave penalized likelihood and its oracle properties , author=. Journal of the American statistical Association , volume=. 2001 , publisher=

    work page 2001

  38. [38]

    Regularized mixture

    Robitzsch, Alexander , journal=. Regularized mixture. 2022 , publisher=

    work page 2022

  39. [39]

    Computational Statistics & Data Analysis , volume=

    Model-based clustering of high-dimensional data: A review , author=. Computational Statistics & Data Analysis , volume=. 2014 , publisher=

    work page 2014

  40. [40]

    Advances in Data Analysis and Classification , volume=

    A LASSO-penalized BIC for mixture model selection , author=. Advances in Data Analysis and Classification , volume=. 2014 , publisher=

    work page 2014

  41. [41]

    Optimization Methods & Software , volume=

    Covariance selection for nonchordal graphs via chordal embedding , author=. Optimization Methods & Software , volume=. 2008 , publisher=

    work page 2008

  42. [42]

    International Journal of Testing , volume=

    Differential item functioning: A mixture distribution conceptualization , author=. International Journal of Testing , volume=. 2002 , publisher=

    work page 2002

  43. [43]

    Applied Psychological Measurement , volume=

    Explanatory secondary dimension modeling of latent differential item functioning , author=. Applied Psychological Measurement , volume=. 2011 , publisher=

    work page 2011

  44. [44]

    Maximum likelihood from incomplete data via the

    Dempster, Arthur P and Laird, Nan M and Rubin, Donald B , journal=. Maximum likelihood from incomplete data via the. 1977 , publisher=

    work page 1977

  45. [45]

    Detection of Differential Item Functioning with Nonlinear Regression: A

    Drabinov. Detection of Differential Item Functioning with Nonlinear Regression: A. Journal of Educational Measurement , volume=. 2017 , publisher=

    work page 2017

  46. [46]

    Journal of Educational Measurement , volume=

    Demonstrating the utility of the standardization approach to assessing unexpected differential item performance on the Scholastic Aptitude Test , author=. Journal of Educational Measurement , volume=. 1986 , publisher=

    work page 1986

  47. [47]

    Handbook of quantitative methods for detecting cheating on tests , pages=

    Detecting preknowledge and item compromise: Understanding the status quo , author=. Handbook of quantitative methods for detecting cheating on tests , pages=. 2016 , publisher=

    work page 2016

  48. [48]

    2013 , publisher=

    Item Response Theory , author=. 2013 , publisher=

    work page 2013

  49. [49]

    Effects of amount of

    Fidalgo, AM and Mellenbergh, Gideon J and Mu. Effects of amount of. Methods of Psychological Research Online , volume=. 2000 , publisher=

    work page 2000

  50. [50]

    Applied Psychological Measurement , volume=

    IRT models for ability-based guessing , author=. Applied Psychological Measurement , volume=. 2006 , publisher=

    work page 2006

  51. [51]

    Educational and Psychological Measurement , volume=

    Investigation of specific learning disability and testing accommodations based differential item functioning using a multilevel multidimensional mixture item response theory model , author=. Educational and Psychological Measurement , volume=. 2013 , publisher=

    work page 2013

  52. [52]

    2010 , publisher=

    Bayesian Item Response Modeling: Theory and Applications , author=. 2010 , publisher=

    work page 2010

  53. [53]

    2010 , publisher=

    Frederickx, Sofie and Tuerlinckx, Francis and De Boeck, Paul and Magis, David , journal=. 2010 , publisher=

    work page 2010

  54. [54]

    Biostatistics , volume=

    Sparse inverse covariance estimation with the graphical lasso , author=. Biostatistics , volume=. 2008 , publisher=

    work page 2008

  55. [55]

    2009 , publisher=

    The Elements of Statistical Learning: Data Mining, Inference, and Prediction , author=. 2009 , publisher=

    work page 2009

  56. [56]

    1993 , publisher=

    Differential Item Functioning , author=. 1993 , publisher=

    work page 1993

  57. [57]

    Differential item functioning and the

    Holland, Paul W and Thayer, Dorothy T , journal=. Differential item functioning and the. 1986 , publisher=

    work page 1986

  58. [58]

    International Journal of Selection and Assessment , volume=

    Determinants, detection and amelioration of adverse impact in personnel selection procedures: Issues, evidence and lessons learned , author=. International Journal of Selection and Assessment , volume=. 2001 , publisher=

    work page 2001

  59. [59]

    Journal of the American statistical Association , volume=

    Estimation of a model with multiple indicators and multiple causes of a single latent variable , author=. Journal of the American statistical Association , volume=. 1975 , publisher=

    work page 1975

  60. [60]

    Journal of Educational Measurement , volume=

    Detection of differential item functioning in multiple groups , author=. Journal of Educational Measurement , volume=. 1995 , publisher=

    work page 1995

  61. [61]

    Anchor selection strategies for

    Kopf, Julia and Zeileis, Achim and Strobl, Carolin , journal=. Anchor selection strategies for. 2015 , publisher=

    work page 2015

  62. [62]

    A framework for anchor methods and an iterative forward approach for

    Kopf, Julia and Zeileis, Achim and Strobl, Carolin , journal=. A framework for anchor methods and an iterative forward approach for. 2015 , publisher=

    work page 2015

  63. [63]

    SIAM Journal on Optimization , volume=

    Proximal Newton-type methods for minimizing composite functions , author=. SIAM Journal on Optimization , volume=. 2014 , publisher=

    work page 2014

  64. [64]

    Multivariate Behavioral Research , volume=

    Detecting social desirability bias using factor mixture models , author=. Multivariate Behavioral Research , volume=. 2010 , publisher=

    work page 2010

  65. [65]

    1980 , publisher=

    Applications of Item Response Theory to Practical Testing Problems , author=. 1980 , publisher=

    work page 1980

  66. [66]

    Lord , title =

    Frederic M. Lord , title =. Basic problems in cross-cultural psychology , editor =. 1977 , publisher =

    work page 1977

  67. [67]

    Multivariate Behavioral Research , volume=

    Improvement in detection of differential item functioning using a mixture item response theory model , author=. Multivariate Behavioral Research , volume=. 2010 , publisher=

    work page 2010

  68. [68]

    Journal of the National Cancer Institute , volume=

    Statistical aspects of the analysis of data from retrospective studies of disease , author=. Journal of the National Cancer Institute , volume=. 1959 , publisher=

    work page 1959

  69. [69]

    Journal of Educational and Behavioral Statistics , volume=

    Detection of differential item functioning using the lasso approach , author=. Journal of Educational and Behavioral Statistics , volume=. 2015 , publisher=

    work page 2015

  70. [70]

    Applied Psychological Measurement , volume=

    A Bayesian method for the detection of item preknowledge in computerized adaptive testing , author=. Applied Psychological Measurement , volume=. 2003 , publisher=

    work page 2003

  71. [71]

    High-dimensional graphs and variable selection with the lasso , author=

    work page

  72. [72]

    Applied Psychological Measurement , volume=

    A mixture Rasch model with item response time components , author=. Applied Psychological Measurement , volume=. 2010 , publisher=

    work page 2010

  73. [73]

    2012 , publisher=

    Statistical approaches to measurement invariance , author=. 2012 , publisher=

    work page 2012

  74. [74]

    Psychometrika , volume=

    Modeling item responses when different subjects employ different solution strategies , author=. Psychometrika , volume=. 1990 , publisher=

    work page 1990

  75. [75]

    Journal of Educational Statistics , volume=

    A method for studying the homogeneity of test items with respect to other relevant variables , author=. Journal of Educational Statistics , volume=. 1985 , publisher=

    work page 1985

  76. [76]

    Psychometrika , volume=

    Latent variable modeling in heterogeneous populations , author=. Psychometrika , volume=. 1989 , publisher=

    work page 1989

  77. [77]

    Some uses of structural equation modeling in validity studies: Extending

    Muth. Some uses of structural equation modeling in validity studies: Extending. Test validity , pages=. 2013 , publisher=

    work page 2013

  78. [78]

    Multiple group

    Muthen, Bengt and Lehman, James , journal=. Multiple group. 1985 , publisher=

    work page 1985

  79. [79]

    The Annals of Statistics , pages=

    Asymptotic properties of criteria for selection of variables in multiple regression , author=. The Annals of Statistics , pages=. 1984 , publisher=

    work page 1984

  80. [80]

    Handbook of quantitative methods for detecting cheating on tests , pages=

    Detecting candidate preknowledge and compromised content using differential person and item functioning , author=. Handbook of quantitative methods for detecting cheating on tests , pages=. 2016 , publisher=

    work page 2016

Showing first 80 references.