Studied by 5 people
Describe Scatter Plot
Form, direction, strength, outliers, unusual patterns, context
r
-1 to 1, tells strength and direction of relationship, closer to zero = weaker
r^2
0 to 1, tells how good of a job the model does, the % of the variation in y that is explained by the variation in x, higher = good
S
Standard deviation of residuals, typical error of predicted y-values from the observed y-values.
Heteroskedasicity
variation of the variance
Slope
for every one unit increase in [X-VAR] there is predicted [SLOPE] inc/dec in [Y-VAR]
Slope Equation
b = (r) sy/sx
y-int equation
(y mean) - b(x mean)
Effect of adding or subtracting a constant
adds/subtracts a from mean/median/quartiles/percentiles, does not change distribution shape or range/IQR/standard deviation
Effect of multiplying or dividing by a constant
multiplies/divides mean/median/quartiles/percentiles, by b, multiplies/divides range/IQR/std. dev., does nt change snape
