University:
California State University, Northridge
Course:
PSY 427 | Introduction to Psychological Testing
Academic year:

2024
Views:

233

Pages:

12
Author:

reannagn8t

Applications  Ordered Polytomous Items  IRT models exist for data that are not dichotomously scored  With dichotomous items there is a single difficulty (location) that indicates the threshold at which the probability switches from favoring one choice to favoring the other  With polytomous items a separate difficulty exists as thresholds between each sets of ordered categories 1 Category thresholds Probability of Endorsing 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -3 -2 -1 0 Latent Trait 1 2 3 1 Item 1 Probability of Endorsing 0.9 Category thresholds 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -3 -2 -1 0 Latent Trait 1 2 3 1 Probability of Endorsing 0.9 0.8 Category Thresholds Item 2 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -3 -2 -1 0 Latent Trait 1 2 3 Applications  Differential Item Functioning  How can age groups, genders, cultures, ethnic groups,      and socioecomonic backgrounds be meaningfully compared? Can be a research goal as opposed to just a test of an assumption Test equivalency of test items translated into multiple languages Test items influenced by cultural differences Test for intelligence items that gender biased Test for age differences in response to personality items “Don’t care about life” Applications  Scaling individuals for further analysis  We often collect data in multifaceted forms (e.g. multi-items surveys) and then collapse them into a single raw score  IRT based scores represent an optimal scaling of individuals on the trait  Most sophisticated analyses require at-least interval level measurement and IRT scores are closer to interval level than raw scores  Using scaled scores as opposed to raw scores has been shown to reduce spurious results Applications  Scale Construction and Modification  The focus is changing from creating fixed length, paper/pencil tests to creating a “universe” of items with known IRF’s that can be used interchangeably  Scales are being designed based around IRT properties  Pre-existing scales that were developed using CTT are being “revamped” using IRT Applications  Computer Adaptive Testing (CAT)  As an extension of the previous slide, once a “universe” (i.e. test bank) of items with known IRFs is created they can be used to measure traits in a computer adaptive form  An item is given to the participant (usually easy to moderate difficulty) and their answer allows their trait score to be estimated, so that the next item is chosen to target that trait level  After the second item is answered their trait score is re-estimated, etc. Applications  Computer Adaptive Testing (CAT)  CA tests are at least twice as efficient as their paper and pencil counterparts with no loss of precision  Primary testing approach used by ETS  Adaptive form of the Headache Impact Survey outperformed the P and P counterpart in reducing patient burden, tracking change and in reliability and validity (Ware et al., 2003) Applications  Test Equating  Participants that have taken different tests measuring the same construct (e.g. Beck depression vs. CESD), but both have items with known IRFS, can be placed on the same scale and compared or scored equivalently  Equating across grades on math ability  Equating across years for placement or admissions tests