NAME______PS 399

Spring 1996

Dr. Nichols

MID-TERM #2

I. Multiple Choice: Please select the best response for the following questions (3 points each).

A. Experiments

1. Which of the following is primarily considered a threat to the external validity of an experiment?

_____ (A) testing effects

_____ (B) representativeness of subject pool

_____ (C) statistical regression

_____ (D) history effects

_____ (E) experimental setting

_____ (F) both A and C

_____ (G) both B and E

_____ (H) both C and D

2. Which of the following is primarily considered a threat to the internal validity of an experiment?

_____ (A) instrument decay

_____ (B) representativeness of subject pool

_____ (C) maturation

_____ (D) history effects

_____ (E) experimental setting

_____ (F) A and C

_____ (G) A, B and D

_____ (H) A, C and D

3. “Statistical regression” effects refer to:

_____ (A) real-world events that occur between the pre-test and post-test

_____ (B) random fluctuation above and below a subject’s “normal” score

_____ (C) change in the instrument used to measure the dependent variable

_____ (D) subjects dropping out of the experiment

_____ (E) none of the above

4. Which of the following is NOT a method for placing subjects into experimental and control groups?

_____ (A) selection

_____ (B) random assignment

_____ (C) precision matching

_____ (D) frequency distribution control

5. A simple post-test experimental design has which advantage over a pre-test/post-test design?

_____ (A) enables a researcher to assess experimental and control group equivalency prior to the treatment

_____ (B) precludes the possibility of testing effects

_____ (C) permits a researcher to take into account possible testing effects

_____ (D) allows a researcher to examine possible history effects

_____ (E) all of the above

_____ (F) both A and C

6. In their study of attack advertising and voter turnout, Ansolabehere et al. created an experimental

setting that they hoped was a fair replication of a natural TV viewing situation. They did so in order to:

_____ (A) reduce the likelihood of instrument decay

_____ (B) lower the possibility of statistical regression

_____ (C) strengthen the external validity of their experiment

_____ (D) rule out the possibility of maturation effects

_____ (E) both A and C

_____ (F) both B and C

7. A Solomon four-group design enables a researcher to take into account:

_____ (A) testing effects

_____ (B) differences between experimental and control groups

_____ (C) any possible interaction between the pre-test and the treatment

_____ (D) all of the above

_____ (E) A and B only

_____ (F) A and C only

8. A multigroup experimental design is primarily intended to allow a researcher to assess:

_____ (A) the internal validity of the experiment

_____ (B) the possible effects of manipulating the independent variable several different ways

_____ (C) statistical regression effects

_____ (D) the external validity of the experiment

_____ (E) control group and experimental group equivalency

9. A factorial design primarily enables a researcher to:

_____ (A) determine the effects of several independent variables, alone and in combination

_____ (B) control for the possibility of non-equivalent experimental and control groups

_____ (C) enhance the external validity of the experiment

_____ (D) assess the impact of a single independent variable

_____ (E) both A and C

_____ (F) both B and D

10. Field experiments differ from laboratory experiments in that:

_____ (A) field experiments give the researcher greater control over the experimental setting

_____ (B) field experiments provide greater internal validity

_____ (C) field experiments provide greater external validity

_____ (D) field experiments are less likely to raise ethical concerns

_____ (E) both A and B

_____ (F) both C and D

B. Surveys and sampling

11. A rolling cross-sectional survey:

_____ (A) examines the same respondents at multiple time points

_____ (B) is based on a non-probability sample

_____ (C) is based on a probability sample

_____ (D) examines respondents at a single time point

_____ (E) both A and B

_____ (F) both C and D

_____ (G) both A and C

12. A panel study:

_____ (A) enhances the internal validity of the survey

_____ (B) examines the same respondents at multiple time points

_____ (C) requires a probability sample

_____ (D) allows a researcher to assess real change over time in respondent attitudes

_____ (E) both A and B

_____ (F) both B and C

_____ (G) both B and D

13. A cross-sectional survey is:

_____ (A) a survey of different respondents at multiple time points

_____ (B) essentially the same as a panel study

_____ (C) subject to problems of panel mortality

_____ (D) good for overcoming problems of statistical regression

_____ (E) both A and D

_____ (F) both B and D

_____ (G) A, C, and D

_____ (H) none of the above

14. A quota sample is an example of a:

_____ (A) panel study design

_____ (B) probability sample

_____ (C) random sample

_____ (D) non-probability sample

_____ (E) rolling cross-section

15. Which of the following is NOT a probability sample?

_____ (A) a cluster sample

_____ (B) a snowball sample

_____ (C) a stratified sample

_____ (D) a systematic sample

_____ (E) a quota sample

_____ (F) both A and B

_____ (G) both B and C

_____ (H) both A and C

_____ (I) both B and E

16. A stratified sample is useful if:

_____ (A) the researcher wants to oversample certain groups in the population

_____ (B) no comprehensive sampling frame exists

_____ (C) the researcher wants to minimize sampling error by minimizing the number of sample stages

_____ (D) the researcher wants to ensure the inclusion of the proper proportions of certain characteristics

in the sample

_____ (E) all of the above

_____ (F) both A and B

_____ (G) both A and C

17. Consider the following survey question: “Do you agree that abortion is murder and that all federal

funding of abortions should cease?” This question is problematic primarily because it is:

_____ (A) double-barreled

_____ (B) vague

_____ (C) too technical

_____ (D) all of the above

_____ (E) both A and B

_____ (F) both A and C

18. One advantage of open-ended questions (as opposed to closed-ended) is that the:

_____ (A) answers are easy to compare, since they fall into pre-determined categories

_____ (B) answers take very little time

_____ (C) answers do not force a respondent to choose answer category that may not represent his/her

position

_____ (D) answers encourage replies to questions about sensitive or personal topics

_____ (E) both A and B

_____ (F) both B and C

_____ (G) both C and D

19. Response rates for mail surveys are:

_____ (A) lower than for any other type of survey

_____ (B) higher than phone survey response rates, because of the latter’s problems with answering

machines

_____ (C) higher than personal interview response rates, because mail contact is easier than in-person

contact

_____ (D) higher than other survey methods, thanks to incentive techniques such as including money and small gifts with the survey

_____ (E) B and C

_____ (F) C and D

_____ (G) B, C, and D

20. A disadvantage of in-person survey interviews is:

_____ (A) the generally poor quality of the interview

_____ (B) the high likelihood that the respondent will end the interview before it is completed

_____ (C) the expense

_____ (D) the possibility that the respondent will not be truthful when asked face-to-face about sensitive subjects

_____ (E) both A and B

_____ (F) both C and D

_____ (G) both B and C

II. Essays: Please answer BOTH of the following questions (20 points each). Feel free to use the back of this page and the extra page, as needed.

21. We noted in class that experiments are “interventionist” -- that they “intrude” on the natural world. In what ways do they do so, and why?

22. The underlying assumption of survey research is that the best way to find out something is to simply ask. Discuss some problems related to the psychology of survey response that call this assumption into question.