Questions about Basic Regression Diagnostics (continued).
(written at Notre Dame to cover material in Chapter 10 of SAS System for Elementary Statistical Analysis)
Click on the answer you think is most nearly correct.
1. The following type of plot does not help you detect a missing variable, the need for a quadratic term, or outliers in the data:
A plot of the predicted variable against an independent variable. A plot of the residuals against a time sequence variable. A plot of the predicted variables against the residuals. A plot of the studentized residuals against observation number. A plot of an independent variable against the residuals.
2. In SAS the following statement is true:
The R option provides a subset of the items printed out as a result of the P option. The CLI option provides a subset of the items printed as a result of the R option. The P option and the R option are just two ways of getting exactly the same items printed. The CLM option provides a subset of the items printed as a result of the P options. The P option provides a subset of the items printed as a result of the R option.
3. An observation may be considered a probable outlier if and only if:
It is highly likely to occur by chance. Its studentized residual is greater than one in absolute value. Its studentized residual is greater than two in absolute value. Its studentized residual is greater than three in absolute value. Its studentized residual is greater than four in absolute value.
4. In general a variable in SAS that starts and ends with an underbar (such as _N_ or _K_) that you did not read in or create yourself, is known as:
A sequence variable. A time variable. A trend variable. An automatic variable. A studentized variable.
5. To obtain a studentized residual you need to:
Take the predicted value, subtract its mean and then divide by its standard deviation. Subtract the value of the independent variable from the value of the dependent variable and divide by the standard error of the difference. After subtracting the observed value from the predicted value divide by the standard deviation of the independent variable. First subtract the predicted value from the observed value and then divide that difference by the standard error of that difference. Take the difference between the fitted value of the dependent variable and its predicted value and divide by its standard error.