BESC1433 Lecture 10: BESC1433 Psychological Assessment and Individual Differences- Week 10 Notes
Document Summary
Week 10: measure development: define measurement error and sources of measurement error, standard error of measurement (duck, test development process (cat, item response format (pug, item analysis (parrot, item response theory (irt) (gecko, criterion references tests (doggo) Measurement error: all tests (scales) contain some degree of error, there is always some inaccuracy when measuring something. Observed score = true score + measurement error. X (what we measure) = t (true score) + e (error) Reliability is a way of saying how much reliability you have. Standard error is a standard deviation of the errors. Sem = 2 means: 2 (1. 96) standard deviations on either side. Sem = sd (1 rtt). r = reliability coefficient: sd = standard deviation on the test, think of the case where r = 1. So the se if the reliability is perfectly 1 will be 0. We want the se to be as small as possible: think of the case where r = 0.