Abstract
Provides item exclusion procedure, which is a formal method to test 'Indifference Of iNdicator' (ION). When a latent personality trait-outcome association is assumed, then the association strength should not depend on which subset of indicators (i.e. items) has been chosen to reflect the trait. Personality traits are often measured (reflected) by a sum-score of a certain set of indicators. Item exclusion procedure randomly excludes items from a sum-score and tests, whether the sum-score - outcome correlation changes. ION has been achieved, when any item can be excluded from the sum-score without the sum-score - outcome correlation substantially changing
Original language | English |
---|---|
Publisher | CRAN - The Comprehensive R Archive Network |
Media of output | Online |
Publication status | Published - 2016 |