# Psychology 2800E Lecture Notes - Lecture 4: Analysis Of Variance, Statistical Hypothesis Testing, Internal Validity

12 views5 pages

15 Aug 2016

School

Department

Course

Professor

Inferential Statistics November 6, 2014

The Logic

-Process of thinking through a logical set of steps

>Based on theories about chance

>Follow steps to decide whether our results are “statistically significant”

>Find the probability that we would obtain the same result just by chance

-Involves logical process of using data from a sample to make inferences about some

population

>Characteristics from sample are known

>Characteristics from population are unknown

-Using information from sample to learn about the population and understand what the

data means

IQ Example

-As you get further away from the mean, fewer people fall in the range so it’s harder to

guess the assumption

Steps of Inferential Stats

1. Assume there is no effect (the null hypothesis)

>Assume that nothing is going

>Ex: no difference between groups or conditions

>Ex: IV does not affect DV

>Ex: variables are not correlated

>Ex: Sarah does not have the ability to identify people with high IQ

-Null Hypothesis: Assumes that the data was due to chance

find more resources at oneclass.com

find more resources at oneclass.com

>Due to chance —>nothing is going on with the data

>NO effects, differences, or correlations in data

2. Collect Data

>Ex: We asked the individual that Sarah identified to complete an IQ test and found

he scored 130

3. Calculate the probability of getting the data that you got if the null is actually true

>What is the probability that we see a difference, effect, or correlation by chance

when it doesn’t exist?

>Ex: What is the probability that Sarah could identify someone with a very high IQ by

chance, when in fact, she does not have the ability to do it?

4. Decide whether you should reject or retain the null hypothesis

>Reject the null=pretty sure that the data were not the result of chance

>Retain the null=based on the data, we are not confident enough to reject the null

Probability Level

-What probability level should we use to accept or reject the null?

-Psychology researchers tend to use 5% or less (p < .05) as their alpha level

>if p < .05, then reject the null

>if p < .05, then retain the null

Errors Associated with Inferences

find more resources at oneclass.com

find more resources at oneclass.com