Unit 3: Exploring Relationships Between Variables Flashcards
Describing Bivariate Data
STD
Describe an association: strength, type, direction
Strength
The closer the dots, the stronger the association
Type (Form)
Linear or Curved
Direction
Positive (goes up) or negative (goes down)
Correlation
Strength of the LINEAR relationship
r
correlation coefficient. Closer to 1 or -1 is stronger
r2
coefficient of determination: percent of variability in y that is explained by variations in x.
slope (b1 ) in context
For every increase in one unit of x, there is an average increase/decrease of b1 units of y.
y-intercept (b0) in context
For a zero amount of x, we expect an average of b0 in y.
Residual
Actual - Predicted
Positive Residual
Actual value is above the LSR. The LSR underestimates the value.
Negative Residual
Actual value is under the LSR. LSR overestimates the value.
LSR
Least Squares Regression Line. Sometimes called the Linear Regression or Line of Best Fit
Residual Plot
Always check to see if the LSR is appropriate. Pattern in the residual plot indicated a curve in the data.
Re-expressing Data
Needs to be done if data are curved or one variable (x or y) is skewed. Likely a natural log or square root.
