Sample Questions for Comprehensive Exam
CONCEPTS
Define and give examples of the following items:
|
QUANTITATIVE METHODS Cross-sectional study Longitudinal study Ecological fallacy Tautology Mutually exclusive and collectively exhaustive Sampling frame Sampling unit Open-ended question Closed-ended question Response rate Experimental group Control group Pretest-posttest design Inductive approach Deductive approach Suppressor variable Units of analysis Scale Index Sample selection bias/missing not at random Probability proportionate to size (PPS) sampling Judgment sampling Secondary analysis EPSEM (Equal Probability of Selection Method) Sampling error |
STATISTICS Intervening Variable Causality Research Hypothesis Reliability Validity Inference Correlation Standardized Regression Coefficient Unstandardized Regression Coefficient Skewness R2 Canonical Coefficient Interval-level Variable Confidence Interval ANOVA F-ratio Outlier Psuedo-R2 Dummy Variable Standard Deviation Quantitative vs. Qualitative Alpha Eta2 Spuriousness Interaction multicollinearity |
Sample Questions (Quantitative Methods) for Comprehensive Exam
Short Answers
Q1
Carefully list and explain at least three characteristics of a good questionnaire.
Q2
Define and give examples of questionnaire items that are (a) contingency questions, and (b) double-barreled questions
Q3
Which is better for explaining causality, an experiment, or a survey? Explain. Which provides the researcher with generalizations that are more reliable, an experiment, or a survey? Explain.
Q4
What is snow ball sampling. Describe three scenarios in which snow ball sampling is preferred than other methods and explain why.
Q5
Describe quota sampling, and produce one example in which quota sampling is used.
Q6
Compare and contrast idiographic and nomothetic. Give an example to each approach.
Q7
Discuss two reasons that anonymity is so important in social survey
Q8
Describe distorter variable in elaborate model and produce one example of distorter variable.
Q9
Can linear regression automatically lead to causality detection? Why or why not?
Q10
Why Likert scale enjoys high popularity over years? Give one example of it.
Sample Questions (Statistics) for Comprehensive Exam
Short Answers
Short Answer
1. Discuss the differences between statistical and substantive significance.
2. Outline and discuss the 6 major steps in the hypothesis-testing.
3. Define statistical interaction and give a concrete sociological example.
4. List and discuss in detail the major assumptions of multiple regression.
5. Discuss in detail the criterion for causality. Give examples for each criterion.
6. Under what circumstances might you use:
- Dummy Regression
- Logistic Regression
- Discriminant Function Analysis
7. Discuss the limitations of using correlation as an analytical technique. How does it compare to regression analysis?
8. Give a verbal description of the following unstandardized regression equation, where Y is a 10-point civil liberties index, X1 is the age in years, X2 is education in years, and X3 is a dummy variable for place of residence (1=rural, 0=urban): (6 points)
Y=6.25-.031X1+.123X2-.546X3
9. Briefly, if I have a slope (b) of -3.43, what does this tell me about the relationship between X and Y?
10. What is an outlier and how do you detect it, and how do you resolve it?
Sample Questions (Quantitative Methods) for Comprehensive Exam
Long Answer
Q1-A researcher wants to study undocumented immigrants, but is plagued by the lack of a sampling frame, from which a random choice of respondents can be made. Identify two research methods that the researcher can use to identify respondents for inclusion in this survey. Discuss in detail how the researcher can use the methods proposed to collect data and compare and contrast the two methods in terms of their advantages and disadvantages.
Q2-A researcher wants to draw a random sample from Prairie Grove to investigate how different people manage work/family scheduling conflicts. However, the city does not have a comprehensive residential list (such as Yellow Book), from which the researcher can draw a random sample. Instead, the city manages its residential list according to a hierarchical structure. The city has five districts. Each district has 10 street blocks. Each street block has a list of all households within the street block. Please advise the researcher how he/she can draw a random sample from this hierarchical structure. In addition, discuss the advantages and disadvantages of your research proposal.
Q3-A research team is conducting an experiment on a new drug to treat diabetes. They randomly selected two groups. In the experimental group, which consists of mostly women in their 20s and 30s, they administered the drug. In the control group, which consists of mostly men in their 70s and 80s, they administered a placebo. The results show that diabetes symptoms were significantly reduced for the experimental group compared to the control group. The tentative conclusion is that the drug works to relieve the diabetes. What are some of the problems with this research design?
Q4-A researcher wants to study the social and religious attitudes of persons who attend Baptist churches. A Baptist religious association presents him a list of 500 churches. Each church maintains their own roster, with the size ranging from 50 to 1000 goers. The research wants to sample 200 respondents. Explain how he/she can use PPS (probability proportionate to size) to sample his respondents.
Q5-A researcher randomly selects two groups, consisting of 10 people each. He/she subjects the experimental group to a daily physical exercise lasting for about 2 hours. For the control group, no physical exercise is required. The experiment lasts for 6 months. The researcher then measures the average weight for the experimental and the control group. He/she found that the average weight for the experimental group drops, while the average weight for the control group remains unchanged. He/she concludes that physical exercise reduces weight. Please identify at least three factors that confound the researcher’s finding.
Q6-The United States continues to experience a growing problem with identity theft. This type of crime is so new that little research has been conducted on this criminological problem. Consequently, even the most basic research questions need to be answered. What are the demographic characteristics of persons committing these crimes? Where are they most likely to occur? Who are the most likely victims? What modus operandi do these offenders use in committing these crimes?
Devise a research plan to help you begin to answer these questions. Describe the method(s) you would use, how you would draw a sample, and who would be included, and the method you would use to elicit answers to the basic questions in paragraph one above.
Sample Questions (Statistics) for Comprehensive Exam
Long Answer
1. Select any substantive sociological problem/issue. Discuss in detail its theoretical foundation and hypotheses structure. What are the advantages and disadvantages of examining this issue/problem using qualitative versus quantitative methods.
2. Discuss the differences between unstandardized and standardized regression coefficients. When should you use them and their weaknesses.
3. Table 2 is part of some regression output generated by SPSSPC. Answer the following questions pertaining to the output and the regression technique that was used:
- Write the full prediction equation and interpret the unstandardized regression coefficients.
- To what test does the F value refer? Interpret the results.
- If we were setting the alpha level at .01, what kinds of conclusions might you reach about the variables in this equation?
- Calculate the standardized regression coefficients and substantively interpret them.
TABLE 2
MULTIPLE REGRESSION OUTPUT
|
Mean |
Standard Dev. |
Cases |
Label | |
|
MEMNUM |
1.87 |
1.87 |
1467 |
Number of Affiliations |
|
CHILDS |
2.05 |
1.89 |
1468 |
Number of Children |
|
EDUC |
12.37 |
3.18 |
1469 |
Years of Education |
|
AGE |
45.43 |
17.80 |
1463 |
Age of Respondent |
|
PSIZE |
2.51 |
1.50 |
1470 |
Pop Size of Place |
|
Correlations |
|||||
|
MEMNUM |
CHILDS |
EDUC |
AGE |
PSIZE | |
|
MEMNUM |
1.000 |
||||
|
CHILDS |
-.064 |
||||
|
EDUC |
.332 |
-.241 |
|||
|
AGE |
-.033 |
.339 |
-.347 |
||
|
PSIZE |
-.017 |
-.093 |
.081 |
-.054 |
Regression Output
Equation Number 1. Dependent Variable..MEMNUM Number of Affiliations Belong To
Block Number 1. Method: Enter CHILDS EDUC AGE PSIZE
Variables Entered on Step
- PSIZE = Pop Size of Place Respondent Lives (In 10,000's)
- AGE = Age of Respondent in Years
- CHILDS = Number of Children
- EDUC = Education Level in Years
- Multiple R .34637
- R Square .11997
- R Square Change .11997
- Adjusted R Square .11755
- F Change 49.62211
- Standard Error 1.75123
- Signif. F Change .00000
- F= 49.622
- Signif. of F= .0000
| Variable | b | SE b | β | SE Beta | Corr. | Part Corr |
| PSIZE | -.053 | .031 | .025 | -.017 | -.043 | |
| AGE | .010 | .003 | .027 | -.033 | .086 | |
| CHILDS | -.012 | .026 | .026 | -.064 | -.012 | |
| EDUC | .215 | . 016 | .027 | .332 | .339 |
(Constant) -1.087 .278
4. A sociologist wants to compare the average (mean) earned income among different ethnic groups. For starters, they use three ethnic groups and randomly draw four subjects from each group. The partial ANOVA results are in the table below.
Source Sum of Squares d.f. Mean Squares F Prob F
Between 72
Within
Total 153
- Fill in the Anova Table
- Test the appropriate hypothesis using the six steps of hypothesis-testing (P<.05) and interpret your results.
5. Using just the information contained below, draw a causal model, estimating all of the path coefficients. Construct a table that shows both the direct and indirect effects. Be sure to estimate the residual paths. Decompose the appropriate correlations showing total and residual effects.
Interpret your model as completely as possible!! BE SURE TO SHOW ALL YOUR WORK. Use the following hypotheses to construct the diagram:
- Increase in X1 leads to a decrease in X2
- Increase in X1 leads to a decrease in X3
- Increase in X2 leads to an increase in X3
|
X1 |
X2 |
X3 | |
|
X1 |
--- |
-.313 |
-.508 |
|
X2 |
--- |
--- |
.307 |
|
X3 |
--- | ||
|
Mean |
5.22 |
1.97 |
24.69 |
|
SD |
3.411 |
1.630 |
11.490 |