Patricia Martinkova

Selected reseach projects

Estimation of psychometric measures as part of admission test development
This is a current junior research grant funded by the Czech Science Foundation. Project aims to describe the current status of admission process to Czech colleges and universities and to prepare a methodology for development of standardized admission tests. Project covers the whole cycle of test development, however, special attention is given to psychometric tools needed for student scoring, item calibration, detection of improper items, estimation of reliability and validity in complex design and student scoring and item selection in automatized adaptive testing. Theoretical results are implemented and demonstrated on real admission data.

Applicants at the Doorstep
This IES project of Center for Education and Data Research analyzes the relationship between measures of teacher applicants (both subjective and objective) and subsequent teacher performance, teacher absences, and teacher attrition in Spokane Public Schools. In this project, I focus on estimation of inter-rater and intra-rater reliability of ratings of applicants (used in the hiring process for selection of candidates invited to interviews) and on its impact on predictive power of the ratings. See the working paper here.

Defining and Assessing the Core Principles for Undergaduate Physiology
In this NSF project of Edmonds Community College I analyze properties of newly designed homeostasis inventory and its items across different groups of undergraduate students.

Admission Tests to School of Medicine at Charles University
Many Czech colleges and universities run their own admission tests. Increased interest in educational testing has led to new analyses of the tests' psychometric properties, the development of more sophisticated item banks and the publishing of books on test development methodology.
In this project I analyze data from admission tests and study outcomes of students at the School of Medicine, Charles University in Prague. I co-authored 5 reports, a paper and an automation of IRT analysis and item calibration using free statistical software R. I also co-authored a book on Student Assessment in Medical Education and delivered presentations on Test Development and Analysis as part of workshops for faculties at Schools of Medicine.

Reliability of Measurements
This study was motivated by a debate in the Czech Republic on the requirement to report Cronbach's alpha, a measure of reliability, for all higher education admissions tests. The study first aimed to explain the statistical groundings for Cronbach's alpha, to stress its assumptions and to research alternative reliability estimates in more complicated designs.
As a secondary outcome, a new estimate of reliability, logistic alpha, was proposed for binary items. The simulation study was performed to compare the logistic alpha with existing estimators of reliability in IRT models. The results were published in papers (here and here) and presented at IMPS and ISCB conferences.

Psychometric Properties of a Set of Clinical Tests Used for MS Patients
The purpose of this study was to develop and validate a battery of clinical tests that can be used to systematically evaluate clinical features of multiple sclerosis treatable by physiotherapy. In this project, I analyzed psychometric properties of the proposed set of clinical tests, co-authored a paper and presentations.

European Survey on Methods Used in Physiotherapy in MS Patients
The aim of this study is to document organizational and non-clinical aspects of physical therapy in multiple sclerosis across Europe. The survey has so far involved 65 centers from 23 countries, and it reveals large differences between European countries in the provision and practice of physical therapy in Multiple Sclerosis. In this project, I provided consultations on the design of a complex questionnaire on Physical rehabilitation in multiple sclerosis and I participate in the data analysis. See also paper here.