Test Bank Ch.4 Elements of Research Design

CHAPTER 4

Elements of Research Design

Multiple Choice

1. Which of the following was not mentioned in the text as a source of topic selection in social research?
   1. theory development
   2. social problems
   3. availability of funding
   4. popular interest
   5. ethical principles
2. To examine whether the unemployment rate is related to rates of welfare and imprisonment, Mark Colvin gathered data from a random sample of 184 urban, industrial counties in the United States. Colvin’s unit of analysis is
   1. the United States.
   2. counties.
   3. a random sample.
   4. welfare recipients.
3. Consider the following hypothesis: Members of the Congress who belong to the same political party as the president are more likely to vote for legislation supported by the president than congressional members who belong to a different party. The unit of analysis implied by this hypothesis is
   1. Congress.
   2. legislative vote.
   3. the political party.
   4. the individual member of Congress.
4. Suppose you develop the following hypothesis: The more cohesive the group, the greater its productivity. To test the hypothesis, you analyze fifteen secretarial pools of six secretaries each in one large insurance company. The unit of analysis in this study would be
   1. the insurance company.
   2. site and time of the study.
   3. the secretarial pools.
   4. individual secretaries.
5. To study whether women are stereotyped on television, Stephen Craig analyzed over 2,000 television commercials. Among the information he coded for each commercial was the gender mix of the characters, the role of the primary visual character, and the type of product advertised. Craig’s unit of analysis is
   1. gender.
   2. gender stereotypes.
   3. women.
   4. television.
   5. television commercials.

1. The ecological fallacy warns researchers to be cautious when drawing conclusions about
   1. individual behavior based on data representing groups or larger social units.
   2. causality based on correlational evidence.
   3. social units—e.g., organizations, cities, nations—based on aggregate data derived from individuals.
   4. strength of association based on evidence of statistical significance.
2. In a sample of corporations, the level of charitable giving increases as the average salary of the corporation increases. It would be an ecological fallacy to conclude that
3. high-paying corporations are more apt to encourage employee giving than low-paying corporations.
4. high-paying corporations give greater amounts than low-paying corporations.
5. charitable giving enhances the economic growth of corporations.
6. employees who are paid more tend to give more to charities.
7. Suppose you found that the incidence of social protest was higher in cities with large union memberships. It would be an ecological fallacy to conclude that
   1. union members are likely to engage in social protest.
   2. unions promote social protest.
   3. cities promote social protest.
   4. social protest encourages unionization.
8. Which of the following research studies would be susceptible to the ecological fallacy?
   1. Observing individuals interacting in a group to determine whether men or women initiate more topic transitions.
   2. Analyzing data from colleges on the percentage of minority students and the average amount of financial aid to determine if minority students are more likely than others to receive financial aid.
   3. Analyzing data from nations on type of government and rate of voluntary association membership to determine if membership rate is higher in democratic than nondemocratic societies.
   4. Interviewing adolescents to see if exposure to violence is related to dropping out of school.
   5. Interviewing respondents to see if the racial composition of a neighborhood affects their preferences about where to live.
9. In which of the following lists does each term denote a variable?
   1. psychology major, sociology major, economics major
   2. race, class rank, age
   3. female, sociology major, sophomore class
   4. family income, sociology major, student
10. In which of the following lists does each term denote a variable?
    1. male, Protestant, year in college
    2. architect, engineer, lawyer
    3. income, Democrat, eighteen years of age
    4. religious affiliation, gender, years of schooling
11. Suppose you hypothesize that among college students women consume alcoholic beverages less often than men. The independent variable is
    1. women students.
    2. college students.
    3. gender.
    4. frequency of alcohol consumption.
12. It is hypothesized that among first-year students, women attend class more frequently than men. Class attendance is a(n) __________ variable, and class standing (first-year, second-year, etc.) is a(n) __________variable.
    1. independent; dependent
    2. independent; control
    3. dependent; independent
    4. dependent; control
13. A study of gender differences in the values of U.S. adolescents found that females were more likely than males to express concern for the well-being of others. This difference was the same across all social class groupings. In this study, the dependent variable is __________, and a control variable is __________.
    1. gender; social class
    2. concern for others; social class
    3. concern for others; gender
    4. social class; gender
14. In a study of the relationship between level of education and support for immigration reform, a person’s race or ethnicity is an __________ variable and their occupational status is an __________ variable.
    1. antecedent; antecedent
    2. antecedent; intervening
    3. intervening; antecedent
    4. intervening; intervening
15. In Broh’s study of the relationship between participation in interscholastic sports and academic achievement, social capital was an __________ variable and parent’s income was an __________ variable.
    1. antecedent; antecedent
    2. antecedent; intervening
    3. intervening; antecedent
    4. intervening; intervening
16. Which of the following statistical properties would not apply to the relationship between the two “qualitative” variables race and political party affiliation?
    1. association
    2. significance
    3. strength
    4. direction (positive or negative)
17. In a recent campus survey, men reported that they watched an average of 104 minutes of television per day, and women reported an average of 79 minutes. This indicates that there is ______ between gender and the number of minutes of TV watched per day.
18. no association
19. an association
20. a linear relationship
21. a curvilinear relationship
22. If older people are less likely to attend church than younger people, the relationship between age and church attendance is
    1. positive.
    2. negative.
    3. curvilinear.
    4. symmetric.
23. Suppose you found a negative correlation between how often respondents report that they attend church and how much money they claimed in charitable contributions on their most recent income tax return. This would indicate that
    1. church attendance and dollar amount of charitable contributions are unrelated.
    2. as church attendance decreases, dollar amount of charitable contributions decreases.
    3. as church attendance increases, dollar amount of charitable contributions decreases.
    4. you made a mistake in your calculation.
24. A correlation coefficient of .10 between SAT score and college GPA is statistically significant at p < .01. This indicates that the association is
    1. weak but significant.
    2. weak and nonsignificant.
    3. strong and significant.
    4. strong but nonsignificant.
25. A correlation coefficient provides information on all but which one of the following properties?
26. statistical significance
27. how changes in the values of one variable are related to changes in the values of another
28. direction of the relationship (positive or negative)
29. strength of association
30. Based on a random sample survey of a college student body, you find that the correlation between the frequency of reported church attendance and the frequency of reported alcohol consumption is −.12. This correlation was significant at p < .01. This means that
    1. the correlation is not significant.
    2. the relationship is spurious.
    3. the more often a student attends church, the more frequently he or she drinks.
    4. there is a weak but significant negative correlation between these variables.
31. Respondents in a campus survey are drawn from a random sample of the entire student body. Therefore, it is appropriate to conduct tests of statistical significance. A significance level of p < .001 for a particular relationship would indicate that the relationship
32. is not significant.
33. can be generalized to other student populations.
34. is likely to exist for the entire student body.
35. easily could have occurred by chance or random processes.
36. Sampson and Laub studied the effect of childhood deviance on adult crime. They found that children who had been committed to a correctional school for persistent delinquency were much more likely to have committed crimes as adults. Which criteria(on) for establishing causality are (is) satisfied in this study?
    1. association
    2. direction of influence
    3. nonspuriousness
    4. all of the above
    5. a and b
37. Tests of statistical significance indicate that which of the following criteria for establishing causality has been met?
    1. association
    2. direction of influence
    3. nonspuriousness
38. In which of the following hypotheses is the independent variable (in other words, the temporal order or direction of influence) most clear?
    1. The more often you interact with someone, the more you like him or her.
    2. Students who have firm career goals achieve higher GPAs than those who do not have firm career goals.
    3. The more often a student drinks, the more likely that he or she is depressed.
    4. The greater the level of parent-child communication, the less likely that a child is delinquent.
    5. Children who are breast-fed tend to have higher IQ scores than those who are not breast-fed.
39. A nonspurious relationship implies that
    1. the independent and dependent variables are interchangeable.
    2. there is a no statistical association between the variables.
    3. there is a positive association between the variables.
    4. there is a negative association between the variables.
    5. there is a causal link between the variables.
40. Data from a U.S. survey showed that boys whose parents were divorced and separated had more behavioral problems than boys living with both biological parents. Evidence that this finding held up when the researchers controlled for social class and race would
    1. clarify the direction of influence.
    2. signify that the relationship is statistically significant.
    3. strengthen the inference that the relationship is nonspurious.
    4. prove that divorce or separation leads to behavioral problems.
41. If an empirical relationship between an independent variable and a dependent variable is maintained when the effects of all antecedent variables are controlled or held constant, the relationship is
    1. spurious.
    2. nonspurious.
    3. positive.
    4. negative.
42. Suppose you find that the more educated people are, the more likely they are to vote. When you analyze the impact of the variable “interest in politics,” you find that education is associated with interest in politics, and people who are interested in politics are more likely to vote. As a (an) __________ variable, interest in politics __________.
    1. antecedent; indicates that the original relationship is spurious
    2. antecedent; explains how education affects likelihood of voting
    3. intervening; indicates that the original relationship is spurious
    4. intervening; explains how education affects likelihood of voting
43. Suppose there is an association between variables X and Y. Showing empirically that a variable intervenes between X and Y
    1. adds to theoretical understanding of the relationship between X and Y.
    2. demonstrates that the relationship between X and Y is spurious.
    3. weakens the inference that a causal relationship exists between X and Y.
    4. weakens the inference that the relationship between X and Y is statistically significant.
44. Several studies have found that students who drink heavily are more likely than other students to miss class, get behind in school work, and perform poorly on examinations. From this finding, we can conclude confidently that
    1. heavy drinking has a negative impact on academic performance.
    2. heavy drinking impairs the cognitive tasks required to perform well academically.
    3. poor academic performance causes distress that leads to heavy drinking.
    4. there is an association between alcohol use and academic performance.
45. In a survey of American high schools, it was found that students who attended private schools showed higher academic achievement than students who attended public schools. Which of the following variables could create a spurious correlation between type of school (private versus public) and academic achievement?
    1. parent’s income
    2. school attendance
    3. amount of homework assigned
    4. strictness of school discipline
46. In her analysis of data from a national survey of high school students, Beckett Broh found that students who played interscholastic sports tended to perform better academically than those who did not play; moreover, this association was statistically significant. Based on this information, it would be appropriate to conclude that
    1. playing sports is a cause (among other causes) of academic performance.
    2. there is an association between playing high school sports and academic performance.
    3. playing sports has a positive impact on academic performance.
    4. the relationship between sports participation and academic performance is nonspurious.
47. Morris (Box 4.1) found a higher incidence of heart disease among drivers than among conductors of London’s double-decker buses. Which of the following variables may have created a spurious association between job type and heart disease?
    1. exercise
    2. job stress
    3. salary or income
    4. age
48. According to the text, good research questions should be
    1. researchable and interesting.
    2. broad and unanswered.
    3. relevant to current social issues.
    4. value neutral.
49. To state that a good research question should be “interesting” means that it should
    1. be of interest to the widest possible audience.
    2. be of strong personal interest to the researcher.
    3. extend previous research in new directions.
    4. allow for the possibility of discovering unanticipated empirical patterns.
50. In her review of the literature on sports participation and grades, Broh found that previous research
    1. failed to examine the impact of sports participation on grades.
    2. showed that high school athletes got lower grades, on average, than non-athletes.
    3. showed that students who participated in non-sports extracurricular activities got higher grades than other students.
    4. had not adequately examined theoretical explanations of how sports participation affects grades.
51. Which of the following is the best statement of a hypothesis?
    1. Social class is related to party affiliation.
    2. Among the elderly, marital status is related to happiness.
    3. Protestants tend to be Republicans.
    4. Women are more likely to oppose pornography than are men.
    5. Political conservatives support authoritarian leaders.
52. Which of the following is not an appropriate statement of a hypothesis?
    1. Does place of residence—urban versus rural—affect tolerance?
    2. Rural residents are less tolerant than urban residents.
    3. As the size of one’s community increases, one’s level of tolerance increases.
    4. If someone resides in a rural area, then he or she tends to be intolerant; if someone resides in an urban area, then he or she tends to be tolerant.
53. Which of the following is the least satisfactory statement of a hypothesis?
    1. As GNP increases, social inequality increases.
    2. Men are less religious than women.
    3. Happiness is related to intelligence.
    4. Women tend to receive less pay than men for the same work.
    5. The greater the social distance between criminal and victim, the more severe the punishment.
54. Concerned that the faculty may be underestimating how much time students spend studying, the student government commissions you to conduct a study that will estimate how much time, on the average, the students at your school study per day. This is an example of a(n)
    1. exploratory study.
    2. descriptive study.
    3. explanatory study.
    4. hypothesis-testing study.
55. Selecting the sampling and measurement techniques is part of what stage of the research process?
    1. data collection
    2. data processing
    3. problem selection and formulation
    4. formulation of the research design
    5. data analysis and interpretation

True and False

T F 1. Personal interest is a poor basis for topic selection.

T F 2. In a study comparing acceptance rates with graduate rates at numerous colleges, the unit of analysis is individual students.

T F 3. Emile Durkheim may have committed an ecological fallacy in his classic study of suicide by inferring individual propensities to suicide from group suicide rates.

T F 4. “Divorced,” “male,” “and “Republican” are examples of variables.

T F 5. Independent variable is to cause as dependent variable is to effect.

T F 6. An independent variable in one study may be a dependent variable in another study.

T F 7. Extraneous variables may be either controlled or uncontrolled.

T F 8. Including only males in a study controls for the variable “gender.”

T F 9. In a study of the effect of gender on trust, gender would be the control variable.

T F 10. Anticipated relationships guide all research, even research aimed exclusively at exploration or description.

T F 11. A primary objective of research design is to plan a study so that the facts speak for themselves without the need for subjective interpretations.

T F 12. The best control for investigator bias is to try not to anticipate any patterns or relationships among variables.

T F 13. A correlation coefficient indicates whether a relationship is statistically significant.

T F 14. A correlation coefficient indicates the strength of a relationship.

T F 15. A negative correlation between variables means that the variables are not causally related.

T F 16. Statistical significance indicates whether a relationship is spurious or nonspurious.

T F 17. In a study of the impact of gender on aggression, the direction of influence is clear.

T F 18. Either an antecedent or intervening variable may create a spurious relationship.

T F 19. Showing that an intervening variable exists between an independent and dependent variable proves that the relationship is spurious.

T F 20. A causal relationship may be implied from comparatively weak associations.

T F 21. The three criteria for establishing causality are association, direction of influence, and nonspuriousness.

T F 22. Research questions are defined as “interesting” when they are about current social issues.

T F 23. A given hypotheses may be stated properly in several different ways.

T F 24. The three main purposes of social research are discovery, explanation, and invention.

Essay

1. A researcher examining data on schools finds that the higher the average income of the parents whose children attend a school, the higher is the average achievement test score for all students in the school. (a) What is the unit of analysis for these data? (b) Can the researcher conclude from these data that the higher the income of a student’s parents, the better the student’s achievement test score is likely to be? Why or why not?
2. In what sense do anticipated relationships among variables guide research design?
3. Suppose you want to investigate the effect of management leadership style on organizational performance (in terms of, for example, level of profit, market share, and product quality).
   1. What would be the independent variable in your study?
   2. What would be the dependent variable?
   3. What would be the unit of analysis?
4. Explain the difference between the following four properties of relationships: strength, directionality, linearity, and statistical significance. Which properties are revealed by a correlation coefficient?
5. Suppose a researcher gathered data from a sample of children that measured the children’s aggressiveness and the number of brothers and sisters in each child’s family. The data showed that the more siblings a child has, the greater the child’s level of aggressiveness. Realizing that there may be more than one cause of aggressiveness, could you conclude from this finding that number of siblings is a cause of level of aggressiveness? Carefully explain.
6. Rewrite the following statement as a testable hypothesis: Social class is related to political party affiliation.