PSYCH 100A Lecture Notes - Lecture 9: Pooled Variance, Standard Error, Standard Deviation
Document Summary
Two sample: t-test only, independent or dependent. A study was conducted comparing orphans with non-orphans on the basis of the judge"s size of parental figures viewed at a distance. Sample 1 was a group of orphaned children without foster parents. Sample 2 was a group of orphaned children with parents. H0: mu1- mu2 less than or equal to zero. Ha: mu1 - mu2 is greater than zero. Both samples are independent of each other because one does not affect the other. Df(total): df1 + df2 = 124+149 = 273. Pooled variance: 0. 49 (124/273) + 0. 81(149/273) = 0. 66. Calculate pooled variance estimate, based on assuming variance is equal in population 1 and 2 (and normality) Take the square root at the end of the problem. If there are 50% in each group, just add the two variances together and divide by two. Standard error(x1): pooled variance/ n : . 66/125 = . 0053. Standard error(x1-x2) = . 0053 + . 0044 = . 0097.
