SCMA 320 Study Guide - Midterm Guide: Central Limit Theorem, Confidence Interval, Sampling Distribution

24 views3 pages

Document Summary

S = sq. root (x xbar )^2/n-1. Bell curve and unimodal (standard normal mu=0, =1, x=2. Probability on a normal distribution (either using a z-score and the table or excel) Find p(z <= -1. 72) -> z table: 0. 0427. A sampling distribution is made by plotting statistics (p-hats) Purpose: to study how statistics vary from sample to sample. What does unbiased mean: unbiased: if i average all the statistics in the sampling distribution, i get the parameter, small spread (consistent): small spread -> one statistic is more likely to be close to parameter than large spread. Sample range is a biased statistic -> it always underestimates the parameter. How to test if the shape will be normal. Spread -> (p-hat) = as spread goes up, (p-hat) goes down. (sq. root p(1-p)/n: normal if: n*p >= 10 and n(1-p) >= 10. Confidence intervals (what will happen to margin of error and interval when confidence level changes?)