Deck 16: Problems and Prospects: Regression Diagnostics, Advanced Techniques, and Where to Go Now

Question 1

The field of statistics concerned about finding and correcting problems within regression models is called:

A) regression troubleshooting
B) regression diagnostics
C) regression medicine
D) regression health

Answer 1

B

Question 2

Which of the following can offer signs that there is a problematic outlier in your data?

A) the R-squared value
B) the statistical significance
C) the standard deviation
D) all of the above

Answer 2

D

Question 3

You run some examples using your regression results, and they just don't make sense given what you know about the world (i.e. they say that as education rises, income goes down). You are putting your results through the:

A) eyeball test
B) sniff test
C) logic test
D) sense test

Answer 3

B

Question 4

If two of your independent variables are highly correlated with one another, you might have a problem with:

A) nonlinearity
B) causality
C) multicollinearity
D) explained variation

Question 5

Concerned about multicollinearity between two of your independent variables, you look at the Tolerance statistic, and it is .98. You should be:

A) not concerned at all
B) a little concerned
C) somewhat concerned
D) very concerned

Question 6

Concerned about multicollinearity between two of your independent variables, you look at the Tolerance statistic, and it is .25. You should be:

A) not concerned at all
B) a little concerned
C) somewhat concerned
D) very concerned

Question 7

You have identified multicollinearity as a problem in your regression model. Which of the following is NOT a suggestion the textbook makes with regard to what to do?

A) remove one of the involved independent variables from the model
B) combined the involved variables into an index
C) create an interaction effect with the two involved variables
D) all of the above are suggestions the textbook makes.

Question 8

We want to use a satisfaction variable (measured: very satisfied, somewhat satisfied, not very satisfied, not at all satisfied) as our dependent variable. Which technique might we want to use?

A) regular linear regression
B) logistic regression
C) ordered logistic regression
D) multinominal logistic regression

Question 9

We want to use a work status variable (measured: working fulltime, working part-time, unemployed, retired) as our dependent variable. Which technique might we want to use?

A) regular linear regression
B) logistic regression
C) ordered logistic regression
D) multinominal logistic regression

Question 10

Tobit regression is most similar to which kind of regression?

A) regular linear regression
B) logistic regression
C) nonlinear regression
D) standardized regression

Question 11

For which of the following variables might someone suggest using negative binomial regression?

A) number of marriages
B) age at first marriage
C) length of marriage
D) cost of wedding

Question 12

You have data from students, parents, teachers, and school principals. You might want to use which technique?

A) path analysis
B) poisson regression
C) multi-level modeling
D) negative binomial regression

Question 13

The technique of taking many, many samples from within your sample, and calculating statistics from these samples, is called:

A) multisampling
B) multi-level modeling
C) repeated measures
D) bootstrapping

Question 14

Which of these advanced techniques is similar to path analysis?

A) structural equation modeling
B) multi-level modeling
C) hazard modeling
D) runway modeling

Question 15

If your research question involves time, such as figuring out when, if ever, something will happen, you might want to use which advanced technique?

A) structural equation modeling
B) multi-level modeling
C) negative binomial regression
D) hazard modeling

Question 16

Your friend shows you some SPSS output for a regression mode she ran, and you notice that the amount of total variation seems extremely high, and even though none of the slopes are statistically significant, the R-squared value is .25. What might be going on here?

Question 17

Your friend shows you some SPSS output for a regression model he ran, and you happen to notice a Tolerance statistic of .30. Tell your friend what problem he might have and describe the problem.

Question 18

You want to study zoos: each separate animal, the zookeepers in charge of each type of animal, and the head of the zoos. Explain what advanced technique you might want to use.

Question 19

Describe one way structural equation modeling is similar to path analysis, then describe one way it is different.

Question 20

You want to study why and when students drop out of high school. Explain what advanced technique you might want to use.