PSYC 406 Lecture Notes - Lecture 5: Classical Test Theory, Confidence Interval, Observational Error
Document Summary
You can assess how much junk is in a test through reliability. Standard error = expected error of what you"re doing. We have discussed different forms of reliability (internal consistency, parallel forms, test-retest, and inter- rater). Psychology has a lot of measurement error, to the point where you may not be able to measure a construct. In psychology, we have to measure a construct, but we have to also decide what the construct is this creates a lot of difficulty. X = obtained score, t = true score, e = error. We collect x, but we have to guess t and e. Error is random error (classical test theory assumes error is random). When you measure people"s heights, the rubber yard stick will be really long, or really short, or a little short, or a little long it misses in a random fashion. In real life, there is also systematic error.