Bekijk de nieuwe studieresultatenpagina. Weer een stap richting de juiste info op één plek

Welke opleiding volg je?

Colloquium Graduate School of Child Development and Education on Monday 28 November 2022

Laatst gewijzigd op 24-11-2022
The next Graduate School colloquium will be on Monday 28 November. Dr. Niel Smits will present recent research findings (for abstract, see below).
Toon informatie voor jouw opleiding
Nu zie je algemene informatie op deze site. Kies je opleiding om ook informatie te zien die specifiek voor jouw opleiding geldt, zoals deadlines, regelingen en contactgegevens.
Welke opleiding volg je?
28-11-2022 12:00
28-11-2022 12:30


With the aim of promoting academic exchange and cooperation, we would like to invite all (research master) students, researchers and other staff members to attend the colloquium. 

Day:               Monday 28 November 2022
Time:            1pm - 1.30 pm CET
Loction:    REC-B3.09 

We look forward to seeing you at the Graduate School Colloquium!   

With kind regards, also on behalf of Niels Smits, Andries van der Ark, Annette van Maanen, Qingqing Du (q.du@uva.nlExterne link) and Jiajun Mo ( link).    


Measurement versus Prediction in Test Construction
Niels Smits, University of Amsterdam

 Two important test goals are (i) measurement: questionnaires are constructed to assign numerical values that accurately represent the test taker's attribute, and (ii) prediction: the questionnaire is constructed to give an accurate forecast of an external criterion. Construction methods aimed at measurement prescribe that items should be reliable, leading to questionnaires with high inter-item correlations. By contrast, construction methods aimed at prediction prescribe that, in addition to a high correlation with the criterion, items should have low inter-item correlations. The latter approach has often been said to produce a paradox concerning the relation between reliability and validity (Gulliksen, 1950; Lord & Novick 1969, McDonald, 1999), because a well-known result from classical test theory states that the validity of a test with respect to any criterion is bounded above by its reliability, by which it may appear more appropriate to use items with high rather than low inter-item correlations. The first goal of the present paper is to provide insight into this seemingly paradoxical situation and the second goal is to show that there is a trade-off between the measurement and prediction qualities of a questionnaire. To that end, both a theoretical exposition and an empirical illustration are presented. We claim that a trade-off will exist for any pair of test goals, and the implications for the construction of questionnaires are discussed.Graduate School Colloquia for the next academic year will take place on:       

  • 30 January 2023
  • 27 February 2023
  • 27 March 2023

Colloquiumcategorie: scriptiepresentaties