PSYC 2360 Lecture Notes - Lecture 4: Internal Consistency, Theoretical Definition, Observational Error
Document Summary
The extent to which a measure is free from random error. To assess reliability: test-retest extent to which scores on the same measure, administered t two diferent times, are correlated. E. g. reliability of adult iq on wais is . 96. Limitations: reactivity (already know the test the second time), some concepts are not stable over time (life satisfaction) Use tests that are the same activities but diferent slightly o o. Inter-rater reliability all judges agree on the same score. Internal consistency reliability extent to which all the items in a multi-item measure correlate with each other: e. g. self-esteem measure, midterm. 70 is acceptable o: number of items has an inluence on the size of the co-eicient. Inter-observer agreement (behavioural measures) judges agree but are independent: extent to which the ratings of one or more observers correlate with each other. Judges need to be trained and independent o. Skating judges score a bunch of things, intense system to reduce cheating.