PSY-0031 Lecture 8: Hypothesis Testing with Binomial Distribution
15 Feb 2017
School
Department
Course
Professor
Document Summary
In classical hypothesis testing with information (statistics) from a sample. , a hypothesis is a statements about a population tested. There are many ways to test a scientific hypothesis - from here on, hypothesis refers to a specific procedure in classical statistics for this course testing. Null hypothesis (h 0 ) - nothing special is happening. There are no differences between groups, and there is no effect of a psychological treatment/drug. Alternative hypothesis (h 1 ) - something is happening. There are differences between groups, there is an effect of a psychological treatment/drug. First define our null and alternative hypotheses. *this is confusing, but we will address it later. If the null hypothesis is actually true (there is no effect), then two things can happen: We correctly continue to assume the null hypothesis. We incorrectly reject the null hypothesis (false alarm) When the null hypothesis is correct but we reject the null anyway.
