Biology 2244A/B Lecture Notes - Lecture 14: Confidence Interval, Sample Size Determination, Statistical Hypothesis Testing
11 Dec 2013
School
Department
Course
Professor
Document Summary
Independent samples from one population not related to samples from the other. In other words, samples don"t have anything to do with each other: for both sample, number of successes and number of failures is at least five. Notation: p1 population proportion, n1 size of sample from pop. 1: x1 number of successes from pop. 1: p-hat1 sample proportion for pop. Finding x1 and x2: x1 = n1*p-hat1. Hypothesis testing: only test that p1=p2. Pooled estimate of p1 and p2: p-bar = (x1+x2)/(n1+n2) For test statistic for two proportions: testing proportions always use z-scores. Confidence intervals for proportions: estimate of difference between poppulation proportions. Inference about two means independent vs dependent*** Independent if values from one population don"t affect values from another. Dependent if members of one sample can be used to determine members of another: samples are matching pairs. Independent samples: both are srs, both samples large (n1>50 and n2>50) or both samples from normal populations.
