HSS 2381 Lecture Notes - Lecture 5: Human Height, Pooled Variance, Type I And Type Ii Errors
17 Dec 2017
School
Department
Course
Professor
Document Summary
Factors that influence hypothesis test (z-score: difference between sample and pop means, variability of scores (std dev, sample size. Random sampling so sample represents population. Ex; if approaching a group of people, only interview 1 person from the group. Std dev is not changed by treatment. Mean can change, but std dev must stay same (in order for data to be analyzed with z-test) Normal sampling distribution distribution of sample means must be normal. Central limit theorem (as sample size increases, the distribution of sample means becomes closer to the normal distribution) Used when you specify ahead of time that treatment will cause mean to either increase or decrease. Observation (likelihood of result or one more extreme is less than 5%) vs. By increasing sample size, you can have a statistically significant result even when it shouldn"t be. Because sem formula has root of n in denominator. Cohen"s d addresses true effect of treatment.
