week 3 simple linear regression Flashcards
What is r squared?
The coefficient of determination. A measure of how much variability in variable A is shared with variable B.
What is the regression equation?
Defines a line that minimises the residuals between the actual and predicted values in the sample.
What is the intercept?
Where the line crosses the vertical axis; the value of Y when X is zero.
What is the gradient (regression coefficient)?
Determines how steep the slope is when X increases by 1, Y increases by B1.
What is the predictor (x)?
This is a known value, from the x-axis of the graph.
What is the formula for the regression equation?
Outcome = Constant + (unstandardized) regression coefficient*predictor i.
What is the sum of squares?
The difference between the observed data and the mean value of y.
How do we assess the goodness of fit?
R squared = SS little m over SS little t.
What does assessing the goodness of fit tell us?
provide a gauge of substantive size of fit.
What is adjusted R squared?
controls for sampling statistics and any predictors that might correlate a bit with each other (if there are more than one).
What does adjusted R squared represent?
The amount of variance accounted for in the population.
What are the assumptions for simple regression?
data must be interval or ration, the relationship must be linear, there should be no major outliers, Homoscedasticity, and normality of residuals (NOT normality of variables!).
What is homoscedasticity?
The regression line should be an equally good predictor over the whole sample. Line of best fit should not become less clear as the graph moves on.
