4.3
what does R squared measure
percent or proportion of total variation in the response variable is explained by the least squared regression line
Rsquared interpretation is always given in what format
percentage
how to interpret R squared each time
% of the change in "y" can be explained by the least squared regression line - fill in percent and fill in y variable
unexplained percentage in R squared
100%- R squared= the unexplained percentage
total deviation
= unexplained vairation + explained variation or y-ybar=(y-ybar)+ (yhat-ybar)
residual plot is
a scatterplot of the residuals (y axis) against the explanatory variable (x axis)
the coefficient of determination will always be
between 0 and 1
R squared
coefficient of determination
purpose of residual analysis
d/t if linear model appropriate d/t if variance of the residuals is constant check for outliers
interpret residual plot
if shows a discernible pattern then may not be linearly related
If R Squared = 1
line explains 100% of the variation in the response variable
if R squared = 0 this means
line has no explanatory value
how to determine if linear model is appropriate
residual plot
how to draw residual plot in calculator
variables in list format do a scatter plot first stat, calc, #4, l1 and l2, line reg- little r must run lin reg every time prior to residual plot x list- L1, Ylist, 2nd STAT, #7 for RESID zoom, #9