PSYC 3000 Lecture 4: Tests that use Normal sampling distributions
Document Summary
Psyc 3000 tests that use normal sampling distributions. Procedure for hypothesis tests: *diagrams in lecture slide* Hypothesis testing = process of comparing observed (sample) w/ expected (sampling distribution generated under hypothesis) & sampling distribution portrayed in lecture slide as nice, symmetric, normal distributions; in reality, they can have all sorts of properties. Statisticians like normal distributions; makes life simpler if you use them. In every normal distribution, most extreme 5% of values (2-tailed) lie 1. 96 sd or more from mean (*diagram in lecture slide*: true of normal distributions only; not true for other distributions. Most extreme 1% of values in normal distribution lie more than + 2. 58 sd from centre (*diagram in lecture slide*: these #s were called critical points, this sort of knowledge called mechanics of normal distributions. Expressing some point in normal distribution as certain # of sds from mean = same as finding its z-score.