Confidence intervals √
Hypothesis testing √
Correlation & Regression
Q: Given data of two separate variables how can we determine if a relationship exists?
Correlation is a statistical method used to determine whether a relationship between
variables exists.
Regression is a statistical method used to describe the relationship.
Simple VS Multiple relationships:
• Simple relationships have two variables, one dependent on the other one.
o Q: In a staff of sales people, does years of experience affect sales?
Experience ⇒ Sales
Independent (explanatory) variable: Experience
Dependent (response) variable: Sales
o We call a analysis on simple cause and effect relationships a simple
regression.
• Analysis on multiple relationships is called a multiple regression. In multiple
regressions two or more independent variables are used to predict one dependent
variable.
o Study Hours
GPA ⇒ College Success
High School Success
Relationships can be classified into two categories:
• A positive relationship exists when both variables increase at the same time.
o Height and weight have a positive relationship.
• In a negative relationship, as one variable increases, the other one decreases.
o Party hours and GPA have a negative relationship.
Simple relationships have two variables: x and y.
A scatter plot is a graph of the ordered pairs of data (x, y) where x is the independent
variable and y is the dependent variable.
Docsity.com
