Institute of Education

Research & Expertise to Make a Difference in Education & Beyond

Sixth International Summer School on Applied Psychometrics Held at IOE

Between August 5 and 10, the Sixth International Summer School ‘Applied Psychometrics in Psychology and Education’ took place at IOE.  

Held annually since 2014, the Summer School has established itself as a prime global venue for renowned experts in psychometrics and aspiring early-career professionals to network about the latest and most important perspectives across dimensions in the theory and practice of test development.

Setting a new high for attendance by international faculty and students, the School of 2019 brought together over 50 participants from 12 countries, including the EU, the US, Mexico, China, etc. This year, the School offered three discrete tracks that covered topics and areas central to psychometric R&D, such as the Rasch Model, growth modeling, exploratory and confirmatory factor analysis, etc., so that the syllabus best fit the learning and hands-on needs of students with various interests and level of preparation.     

This year marks another important milestone for the School as it was able to further reinforce its status as a leading international forum for world-class experts in test development to share their cutting-edge expertise with junior peers. By providing a collaborative environment of very focused and engaged learning and debate, the School certainly helps expand and unlock new vistas in psychometric R&D while also enabling everyone to grow and develop in many professional ways. 

Isak Froumin 
Head of the HSE Institute of Education

Track One. Foundations and Applications of the Rasch Model

Delivered by Prof. Christine Fox and Prof. Svetlana Beltyukova of the University of Toledo (Ohio), this course basically aimed to introduce beginner psychometricians who are only starting to get the hang of professional tools and methods and are still novice in complex data analysis to the Rasch model for dichotomous data and the Rasch Rating Scale model as the two frameworks that are most commonly used in measurements. Prof. Christine Fox and Prof. Svetlana Beltyukova are recognized among the world’s most prominent authorities on Rasch modeling (Prof. Fox has coauthored one of ‘golden standard’ textbooks on the subject).    

How the track program was designed prompted students to tackle more complex and challenging tasks as they were moving through the coursework day by day, so that even more advanced enrollees were able to expand their theoretical horizons and boost practical skills by developing a deeper understanding of how the essential core mechanisms of tests work.       

Track Two. Modelling the Growth: Vertical Scales and Growth Models; Longitudinal Analysis

The module on growth modelling by Dr. Gary Cook (University of Wisconsin-Madison) covered a variety of concepts and approaches that enable a rater to trace how specific personality features, skills or other developmental attributes of an individual have been evolving over time. Growth models allow benchmarking the scale of this change among several people or across multiple social cohorts, while also providing insights into how the nature of these changes differs for various samples and what factors essentially determine such differences.      

The module ‘Longitudinal Analysis’ by Dr. Theodore Walls (University of Rhode Island) focused on three important areas in the analysis of longitudinal data: simulation and programming, longitudinal designs, and methods for the analysis of change. The main objective was to ensure that students working on research involving longitudinal data develop both the baseline conceptual skills and hands-on experience needed to pursue the design of applied models in their work.

Track Three. Simplifying, Identifying, and Evaluating Dimensions: Exploratory and Confirmatory Factor Analyses

This course by Dr. Gavin Brown, a Professor at the University of Auckland who has supervised over 40 doctoral projects in various areas of psychometrics, aimed to introduce students to principles and practices behind simplifying data into its component dimensions.

Specifically, the program focused on how a defensible number of dimensions can be determined and how that exploration can be tested while emphasizing the importance of testing and comparing multiple plausible alternative models. Also, special attention was paid to identifying and troubleshooting problems that psychometricians are most commonly confronted with in factor analysis, such as negative error variance, etc.