Lecture 17: Correlations – Describing Relationships Between Two Variables 2011, 11, 22
Learning Objectives 1. Construct and interpret scatterplots 2. Understand properties of a correlation: direction, strength, form 3. Compute and interpret Pearson correlation coefficient (r) by hand** 4. Difference between correlation and causation*
Scatterplot Y X Plots one variable against the other A Hours Studied Exam Grade B 5 65 C D E Hours Studied Exam Grade
Direction of a Correlation Positive Correlation: The more hours I studied, the better grade I’ll have Negative Correlation: Number of beer you had the night before midterm and your midterm grade
Form of a CorrelationNon-linear Linear Nonlinear correlation Linear correlation
Strength of a Correlation How spread out the dots around the line Stronger ―――――――――――― Weaker
Strength of a Correlation Perfect “+” Perfect “-” IQ Shoe Size
Pearson’s Correlation Coefficient – Measure the Strength of Correlation Notation Population: ( rho) Sample: r Properties Between -1 to +1 Sign of a correlation coefficient r = 1.0 “perfect positive corr.” r = 0.0 “no relationship” r = -1.0 “perfect negative corr.”
Pearson’s Correlation Coefficient – Strength of a Positive Correlation
Pearson’s Correlation Coefficient – Strength of a Negative Correlation
How to Compute the Pearson correlation coefficient (r)? By hand Step 1: Compute SS X & SS Y Step 2: Sum of the Products (SP) Step 3: Compute r XY
Correlation Doesn’t Equal to Causation Given a correlation of ice cream consumption and cases of drowning, you may speculate “ice cream cause drowning?” r =0.70
Lab 17 Correlation – Recap Scatter plot and fitting line Properties of a correlation Direction (Positive; negative) Form (Linear; nonlinear) Strength (Weak vs. Strong) Compute and interpret Pearson’s correlation coefficient (r) Difference between correlation and causation