# KIN 3982 Lecture Notes - Lecture 8: Pearson Product-Moment Correlation Coefficient, Null Hypothesis, Observational Error

For unlimited access to Class Notes, a Class+ subscription is required.

Statistical Testing

Wednesday, February 27, 2019

2:33 PM

Null Hypothesis:

• The null hypothesis states that there are no differences between groups

o Goal of the study is to reject the null hypothesis

o Do significance testing to determine if you can reject the null hypothesis

o Reject the null hypothesis when p<0.05

o Where does the 0.05 come from?

• If the statistics reveal that there is a 95% chance that the null

hypothesis is false, then p is said to be 0.05

Chi Square: analysis used when dependent variable is categorical

Determine if the distribution to groups is due to changes in the independent

variable or random error

Overview of Data Analysis:

• Tests of relationship (e.g., regression)

o Typically used for surveys

• Statistical tests of difference (e.g., ANVOA, t-test)

o Typically used for experimental designs

Correlation (Typically Surveys)

• Definition: quantify the relationship between two variables

• "r" is Pearson's product moment correlation coefficient

• Values vary between +1 to -1 (0 is no relationship_

• Positive - high score on one variable related to high score on the other

variable

• Negative (inverse relationship)- low score on one variable related to high

score on the other variable or vice versa

• 1 or negative 1 = 1 perfect relationship

• R squared r and you get the percentage variance accounted for by the

relationship

Regression

• Bivariate regression: how well one variable predicts the value of a second

variable (i.e., dependent or criterion variable)