Confidence intervals and Significance Tests for Correlation and Regression (2 of 5)
Regression Slope
Confidence Interval
A confidence interval on the slope of the
regression line can be computed based on the
standard error of the slope (b). The
formula for the standard error of b is:
where s
b is the standard error of the slope,
s
est is the
standard error of the
estimate, and X
M is the mean of the predictor variable. The
formula for the standard error of the estimate is:
for the example data for which S
y = 2.432,

.
The only
remaining term of needed to be computed is:
Σ(X - X
M)²
=
(2-4.833)
2 + (3-4.833)
2 + ... +
(9-4.833)
2 = 30.833.
Therefore,