Chapter 12 Flashcards
T-test for Beta: plan
Linear: scatterplot shows and approximately linear pattern
Independent: assuming there are at least blank available
Normal: the normal probability plot shows and approximately linear line showing normality
Equal variance: there is an even distribution above and below the residual equals zero line on the residual plot
Random: randomly selected
We will use a t-test for Beta
T test for beta: State
Bo= 0, Ba (,not equal) where B is the true slope for the population regression line relating x and y
Alpha equals .05
T-test for Beta: do
Find the DF
Slope minus Beta divided by SE of B = t
TCDF
IF NOT EQUAL TO ONE MINUS ANSWER
T-test for Beta: conclude
There is convincing evidence that the true slope for the regression line relating x and y is positive
T interval for Beta: State
We want to estimate beta, the true slope of the population regression line relating x and y at a — confidence level
T interval for Beta: plan
Same as t-test
We will use a T interval for Beta
T interval for Beta: do
Find DF
Make curve
Invt which gives you t*
b plus or minus t*(SD b)
T interval for Beta: conclude
We are – percent confident that the interval from blank to blank contains the truth slope data relating X and Y
Interpret s
s= _____ tells us that, on average, are predicted values X are off by s
Interpret r-sq
= _____ tells us that the percent variation accounted for between X and Y is _____ %
Interpret standard error of slope
SEb: _____ tells us that if the experiment were repeated many times, the estimated slope would be about _____ off from the population slope, on average
Interpret slope
For every one more unit of why, that X is predicted to increase by x (#)
Interpret y intercept
For 0 units of y, we predict the x to be y (coefficient) units
Power model
Log on both
Exponential model
Log on one
