R²
R²% of the response variable is accounted for by our linear model
Slope: b or b1
Our model predicts the response variable will (increase/decrease) units, on average, for every extra unit of the explanatory variable
Intercept
Our model predicts “a” units on average of the response variable when we have 0 units of the explanatory variable
how to analyze a residuals plot
is the association random?
does the residuals plot display a fairly even spread?
what does not verify a linear model?
R²
as R² goes up…
the better the LSRL will predict ‘y’ for a given ‘x’
what is a residual?
actual y - predicted y
what do you need to say to describe an association
context, units, direction, strength, form (linear, curved), unusual features, SENTENCE OF ASSOCIATION
purpose of correlation: ‘r’
gives direction (sign) and a measure of strength/more objective measure of strength
properties of ‘r’
r is dimensionless, sign of r gives direction, value of r will always lie -1<r<1, non-resistant, LTs do not affect r (but reexpressions do), not affected by which variable on which axis
Standard deviation
The actual values of the response variable vary from our model’s predicted values (with a standard deviation of ____ units) or (with an average deviations of _____ units)