kaefa: kwangwoon automated exploratory factor analysis.

Description Details Author(s) References


kaefa: kwangwoon automated exploratory factor analysis


This library stands for improving research capability to identify unexplained factor structure with complexly cross-classified multilevel structured data in R environment in the automated exploratory factor analysis framework what imports mirt::mirt, mirt::mixedmirt and mirt::mdirt (Chalmers, 2012; Chalmers, 2015).

In practice of applied psychological resarch, so much researcher ignoring the impact of the MMMM (Multiple Membership Multilevel Model) and MM in exploratory factor analysis from unfamiliar with statistical innovations who noted in Sharpe (2013) and Foster, Min, and Zickar (2017).

Moreover, A lot of researcher do not know what is the improper solution in the exploratory factor analysis. That may lead wrong conclusion to research society, The kaefa will filter possible improper solutions during the automated exploratory factor analysis. Filtering the Heywood cased models and fail to pass the second-order test model will help this work. These model will not consider to model selection that they are possible improper solutions.

The kaefa may inspect this issues from the MMMM or MM in statistical learning theory perspectives using model selection criteria like the DIC (Kang, 2008; Kang, Cohen, & Sung, 2009; Jiao, Kamata, Wang, & Jin, 2012; Jiao & Zhang, 2015) with maximising generalisability of the number of factor decisions in every calibration (Kang, 2008; Preacher, Zhang, Kim, & Mels, 2013).

If researcher provide of demographical information in kaefa, kaefa will inspect the optimal number of factor and optimal IRT model, and possible error variances or latent differences from demographic information of respondents.

During the calibration, kaefa consider the these item response models: Rasch, 2PL, 3PL, 3PLu, 4PL, ideal (for dichotomous) nominal, gpcm, graded, grsm, ggum, pcm, rsm, monopoly (for polytomous).

Moreover, factor rotation will decide automatically using Zh for minimizing potential outage of the item as actual criteria. As the default, "bifactorQ","geominQ", "geominT", "bentlerQ", "bentlerT", "oblimin", "oblimax", "simplimax", "tandemII", "tandemI", "entropy", and "quartimax" will try to inspect the optimal structure of actual criteria reflect to the conceptual criterion. It is make a way to increase interpretability of the exploratory factor analysis without the human intervention as objectivity and reproducibility what principles of the science.

After the every n-th calibration, kaefa do the item appropriateness test for check which item contribute to explain conceptual criterion with robustness of aberrant response using Zh, S-X2, PV-Q1. If kaefa find out the improper item, kaefa will exclude the worst one automatically and recalibrating the models until all items are acceptable via statistcal criteria.

This software can be pallelise to multiple computers via LAN even heterogeneous environment, so that applied researchers may expand their research capability more easy with kaefa even data has too complicated structure to calibrate in single machine.

This project started in 2013, and restructured in 2017. Hope to help exploring human behavioural mechanisms in complex contexts.


Seongho Bae [email protected]


Chalmers, R., P. (2012). mirt: A Multidimensional Item Response Theory Package for the R Environment. Journal of Statistical Software, 48(6), 1-29. doi: 10.18637/jss.v048.i06

Chalmers, R. P. (2015). Extended Mixed-Effects Item Response Models with the MH-RM algorithm. Journal of Educational Measurement, 52(2), 200–222. doi: 10.1111/jedm.12072

Foster, G. C., Min, H., & Zickar, M. J. (2017). Review of Item Response Theory Practices in Organizational Research. Organizational Research Methods, 20(3), 465–486. doi: 10.1177/1094428116689708

Jennrich, R. I., & Bentler, P. M. (2011). Exploratory Bi-Factor Analysis. Psychometrika, 76(4), 537–549. doi: 10.1007/s11336-011-9218-4

Jiao, H., Kamata, A., Wang, S., & Jin, Y. (2012). A Multilevel Testlet Model for Dual Local Dependence. Journal of Educational Measurement, 49(1), 82-100. doi: 10.1111/j.1745-3984.2011.00161.x

Jiao, H., & Zhang, Y. (2015). Polytomous multilevel testlet models for testlet-based assessments with complex sampling designs. British Journal of Mathematical and Statistical Psychology, 68(1), 65–83. doi: 10.1111/bmsp.12035

Kang, T. (2008). Application of Statistical Model Selection Methods to Assessing Test Dimensionality. Journal of Educational Evaluation, 21(4), 153–175. Retrieved from http://scholar.dkyobobook.co.kr/searchDetail.laf?barcode=4010022701731

Kang, T., Cohen, A. S., & Sung, H.-J. (2009). Model Selection Indices for Polytomous Items. Applied Psychological Measurement, 33(7), 499–518. doi: 10.1007/s00330-011-2364-3

Mansolf, M., & Reise, S. P. (2016). Exploratory Bifactor Analysis: The Schmid-Leiman Orthogonalization and Jennrich-Bentler Analytic Rotations. Multivariate Behavioral Research, 51(5), 698–717. doi: 10.1080/00273171.2016.1215898

Preacher, K. J., Zhang, G., Kim, C., & Mels, G. (2013). Choosing the optimal number of factors in exploratory factor analysis: A model selection perspective. Multivariate Behavioral Research, 48(1), 28–56. doi: 10.1080/00273171.2012.710386

Reise, S. P., & Waller, N. G. (2009). Item Response Theory and Clinical Measurement. Annual Review of Clinical Psychology, 5(1), 27–48. doi: 10.1146/annurev.clinpsy.032408.153553

Reise, S. P. (2012). The Rediscovery of Bifactor Measurement Models. Multivariate Behavioral Research, 47(5), 667–696. doi: 10.1080/00273171.2012.715555

Sharpe, D. (2013). Why the resistance to statistical innovations? Bridging the communication gap. Psychological Methods, 18(4), 572–582. doi: 10.1037/a0034177

seonghobae/kaefa documentation built on Oct. 9, 2018, 7:34 p.m.