Chapter 4: Scatterplots and Correlation - Detailed Notes
Chapter 4: Scatterplots and Correlation
Explanatory and Response Variables
- Response Variable: Measures the outcome of a study.
- Explanatory Variable: Influences or explains changes in the response variable.
Scatterplots
- The most effective graph for displaying relationships between two quantitative variables.
- Both variables are measured on the same individuals:
- Horizontal Axis (x-axis): Explanatory variable (if applicable);
- Vertical Axis (y-axis): Response variable.
- Each point on the scatterplot represents an individual and its corresponding values for both variables.
Steps to Create a Scatterplot (Four-Step Process):
- STATE: Formulate the research question as a statement about the association between the two variables.
- PLAN: Create the scatterplot according to the guidelines stated above.
- SOLVE: Analyze the scatterplot for any observable relationships.
- CONCLUDE: Draw conclusions based on the analysis (details on this step are explored further in the chapter).
Example of a Scatterplot:
- Comparing the percent of high school graduates taking the SAT to the state’s mean SAT Mathematics score.
- Percent taking SAT: 23, 55, 80, etc.
- Mean SAT Math Scores: 566, 551, 538, etc.
Interpreting Scatterplots
- Analyze patterns and deviations in the scatterplot:
- Overall Pattern: Direction, form, and strength of the relationship.
- Outliers: Individual data points that deviate significantly from the general pattern.
Direction of Association:
- Positive Association:
- Above-average values of one variable accompany above-average values of the other, and vice versa for below-average values.
- Negative Association:
- Above-average values of one variable accompany below-average values of the other (and vice versa).
Adding Categorical Variables to Scatterplots
- Differentiate categories using distinct colors or symbols.
- Example: mean SAT Mathematics score against the percent of graduates taking the SAT, differentiating by regions (Midwest and Northeast).
Measuring Linear Association
- Scatterplots reveal strength, direction, and form of relationships.
- Correlation (r): Measures the direction and strength of the linear relationship between two quantitative variables.
- Calculation involves means and standard deviations of both variables.
- Correlation formula:
- [ r = \frac{1}{n-1} \sum (xi - \bar{x})(yi - \bar{y}) ]
Facts about Correlation
- No distinction between explanatory and response variables in correlation.
- Unitless and invariant to changes in measurement units of the involved variables.
- Positive values indicate a positive association; negative values indicate a negative association.
- Correlation values range between -1 and 1.
Cautions about Correlation:
- Both variables must be quantitative for calculation to be meaningful.
- Correlation does not represent nonlinear relationships accurately.
- Sensitive to outliers, which can heavily influence the correlation coefficient.
- Not a complete summary of two-variable data concisely; therefore, further analysis may be necessary to understand the relationship fully.