PSYC 2040 Study Guide - Final Guide: Null Hypothesis, Sampling Distribution, Repeated Measures Design
2 Dec 2014
School
Department
Course
Professor
Document Summary
The probability of a type one error is equal to the assigned alpha value. The t statistic compares the differences we observed between the two means (numerator) and the difference we would expect due to random sampling assuming the null is true (denominator) Using critical ratios, where one assumes the two population means are equal (the null hypotheses), and it uses the standard error of the sample means. For t-tests, the function changes based on degrees of freedom. The t-distribution is more peaked in the middle and higher in the tails for very small degrees of freedom. The degrees of freedom determine the corresponding t-values. As the degrees of freedom increases, the curve normalizes. Independent samples assumptions: random sampling from two clearly defined populations, populations are normally distributed, variances of the two populations are the same (standard deviations) If the variances are homogeneous, you calculate a pooled variance; if not, separate variance estimates are used.
