Long page
descriptions

Chapter 9   Regression Inference

9.1   Regression models

9.1.1   Normal distribution of response at fixed X

The exercise on this page gives a normal linear regression model and asks for the distribution of the response at a fixed value of the explanatory variable.

9.2   Estimating slope and intercept

9.2.1   Standard error of slope

Any estimator's standard error gives information about its accuracy. The exercise on this page gives the standard error of a least squares line's slope and asks for a roughly calculated interval that is likely to include the underlying model's slope. (T values are not required in this exercise.)

9.2.2   Confidence interval for slope

In the two exercises on this page, confidence intervals for a regression model's slope should be calculated from the least squares slope and its standard error. The second exercise is a little harder -- it asks for various confidence levels.

9.2.3   Influences on accuracy

This exercise asks for the characteristics of a data set that will result in more accurate estimation of the linear model's slope.

9.3   Testing the regression slope

9.3.1   Test for zero slope

This exercise gives the least squares slope and its standard error. The p-value for testing whether the regression slope is zero should be calculated and interpreted.