Patricia Martinkova

Computational Psychometrics Group

Computational Psychometrics Group was initiated at Institute of Computer Science in 2015 through project Estimation of psychometric measures as part of admission test development funded by Czech Science Foundation during years 2015-2017. Project explored statistical and psychometric tools for analysis and improvement of admission process. Project's outcomes included:

Item and Test Analysis with Shiny
- Version 0.1 to 1.2.3 of ShinyItemAnalysis - application for Item and Test Analysis. Current version is available here.
- Version 0.1 to 1.2.3 of ShinyItemAnalysis R package. Current version is available on CRAN.
- Article on ShinyItemAnalysis, published in The R Journal.
- Czech article on ShinyItemAnalysis, published in Testforum.
- Manuscript on semi-real time analysis of admission tests, presented at FedCSIS.
- ShinyItemAnalysis has been presented at useR!2016 (Stanford) and useR!2017 (Brussels), see the video HERE.

Detection of Differential Item Functioning with Non-Linear Regression
- Version 1.1.1 of difNLR R package. Current version is available on CRAN.
- Article on difNLR, published in Journal of Educational Measurement.
- Adela Drabinova was awarded student Travel Award to present poster on difNLR at IMPS2016.

Reliability and Validity in Complex Settings
- Paper on detection of disparities in ratings of internal and external applicants with newly proposed model-based inter-rater reliability, published in PLOS ONE.
- Working paper on estimation of interrater reliability (IRR) in complex settings in teacher hiring.

Development and Validation of tests
- Methods paper on DIF analysis in test development, published in CBE-LSE.
- Article on development and validation of HCI, published in CBE-LSE.

Courses in Psychometrics
- Graduate course NMST570 Selected topics in Psychometrics, taught since 2017.
- NMST571 Seminar in Psychometrics, taught at Charles University since 2019.