Regression Inference
STEP Confidence Interval Significance Test
State We want to estimate the slope 𝛽 of the H 0: 𝛽 = 0
population (or true) regression line relating HA: 𝛽 (<, >, ≠) 0
_____ to _____ with __% confidence. Where 𝛽 is the slope of the population
regression line relating _____ to _____.
𝛼 = ___ (0.05 unless stated otherwise)
Plan Suppose we have n observations on an Suppose we have n observations on an
explanatory variable x and a response explanatory variable x and a response
variable y. Our goal is to study or predict the variable y. Our goal is to study or predict the
behavior of y for given values of x. behavior of y for given values of x.
Check LINER. Check LINER.
Linear: The actual relationship between x and Linear: The actual relationship between x and
y is linear. y is linear.
Independent – Individual observations are Independent – Individual observations are
independent of each other. When sampling independent of each other. When sampling
without replacement, check the 10% without replacement, check the 10%
condition: condition:
Check to make sure that 10 times our sample is Check to make sure that 10 times our sample is
less than the entire population. less than the entire population.
Normal: For any fixed value of x, the response Normal: For any fixed value of x, the response
y varies according to a Normal distribution. y varies according to a Normal distribution.
Equal SD: The standard deviation of y (call it 𝜎) Equal SD: The standard deviation of y (call it 𝜎)
is the same for all values of x. is the same for all values of x.
Random: The data come from a well-designed Random: The data come from a well-designed
random sample or randomized experiment. random sample or randomized experiment.
Because our conditions are met, we will use a Because our conditions are met, we will use a
t-interval for the slope of a regression line 𝛽.
Do On the calculator, choose: On the calculator, choose:
STAT à TESTS à G: LinRegTInt STAT à TESTS à F: LinRegTTest
df = df =
(____, ____) test statistic =
p-value =
(use the equation in the notes if you are given (use the equation in the notes if you are given
a Minitab Output) a Minitab Output)
Conclude We are ___% confident that the interval from Because our P-value = ____ is greater/less than
(____, ____) captures the slope of the the significance level 𝛼 = ___, we (fail to) reject
population (or true) regression line relating H0. There is (not) convincing evidence that
_____ to _____. (alternative hypothesis).
Remember that df = n – 2
Minitab Output
Predictor Coef SE Coef T P
Constant (y-int)
Variable (slope) (SE) (test statistic) (p-value)
S= R-Sq = R-Sq(adj) =