*Please enter 3 lines between the last entry and put in your username using heading 3 then type in your notes.*

riosjosh

Logic of significance testing is the logic of the negative. What is actually tested for a hypothesis is the null hypothesis. The null hypothesis always says no difference or no relationship. So the null hypothesis is what is always tested not the research hypothesis. So if something is really different then you reject the null and retained the alternative or the working hypothesis. If there is a normal distribution with no real significance then you fail to reject the null. Then decide on the probability level which for social scientist is 5%. So alpha ยต is 0.05.

Every time you do an experiment there are four possible outcomes.

HYPOTHOESIS WORLD

IN THE REAL WORLD

Your Decision

Null is True

Alternative is true

Reject null

Type I Error; Alpha

Correct decision

Fail to reject the null

Correct decision

Type II error

You can control the type I error. You just change your alpha. In society we always concentrate on type I errors and almost never type II errors.

## COMM3710 - Fall 2011 - Section 001 - Lecture Notes - 09/20

*Please enter 3 lines between the last entry and put in your username using heading 3 then type in your notes.*## riosjosh

Logic of significance testing is the logic of the negative. What is actually tested for a hypothesis is the null hypothesis. The null hypothesis always says no difference or no relationship. So the null hypothesis is what is always tested not the research hypothesis. So if something is really different then you reject the null and retained the alternative or the working hypothesis. If there is a normal distribution with no real significance then you fail to reject the null. Then decide on the probability level which for social scientist is 5%. So alpha ยต is 0.05.Every time you do an experiment there are four possible outcomes.

You can control the type I error. You just change your alpha. In society we always concentrate on type I errors and almost never type II errors.