RESEARCH METHODS, TEST 2

NOTE: Here is an example of a past Research Methods test, plus some undergraduate tests on the similar subject. They are merely illustrations of the types of tests that I give. The specific questions may already be outdated and obsolete.

(Most recent test example)

RESEARCH METHODS, TEST 2 (100 point test) (2007 class)

1. (20 Points) What is the modal (mode) category for each of the following examples? What is the median category for each of the following examples? List the category names, not the percentages. These are real world data taken from recent national opinion polls (percentages have been adjusted to total 100%, however.).

A. CBS News/New York Times Poll. Sept. 4-9, 2007. N=1,263 adults nationwide. MoE ± 3.

"Which of these comes closest to your view? Abortion should be generally available to those who want it. OR, Abortion should be available, but under stricter limits than it is now. OR, Abortion should not be permitted."

Abortion generally available = 35%

Stricter limits = 40%

Abortion not permitted = 25%

B. Gallup Poll. Nov. 3-5, 2003. N=1,007 adults nationwide. MoE ± 3.
						.
"Taking into consideration both your risk of contracting it and the seriousness of the illness, how worried are you personally about experiencing the following illnesses: very worried, somewhat worried, not too worried, or not worried at all? How about Cancer?"

Very worried = 16%

Somewhat worried = 48%

Not too worried = 19%

Not worried at all = 17%

C. Gallup Poll. Aug. 13-16, 2007. N=1,019 adults nationwide. MoE ± 3.
						.
"Overall, how satisfied are you with the quality of education students receive in kindergarten through grade twelve in the U.S. today? Would you say you are completely satisfied, somewhat satisfied, somewhat dissatisfied or completely dissatisfied?"

Completely satisfied = 7%

Somewhat satisfied = 41%

Somewhat dissatisfied = 36%

Completely dissatisfied = 16%

D. FOX News/Opinion Dynamics Poll. July 17-18, 2007. N=900 registered voters nationwide. MoE ± 3. "How concerned are you that the government is using surveillance technologies like cameras and monitoring of personal information to get private information about you? . . ."

Very concerned = 47%

Somewhat concerned = 30%

Not very concerned = 9%

Not concerned at all = 14%

E. Gallup Poll. Oct. 13-16, 2005. N=1,012 adults nationwide. MoE ± 3.
					.

"Overall, how would you describe the problem of crime in the United States? Is it extremely serious, very serious, moderately serious, not too serious, or not serious at all?"

Extremely serious = 19%

Very serious = 30%

Moderately serious = 47%

Not serious at all = 4%

F. NBC News/Wall Street Journal Poll conducted by the polling organizations of Peter Hart (D) and Neil Newhouse (R). June 8-11, 2007. N=1,008 adults nationwide. MoE ± 3.1.

"How much of an impact has illegal immigration made on your own community where you live: a great deal of impact, quite a bit of impact, just a little impact, or no impact at all?"

Great deal of impact = 26%

Quite a bit of impact = 22%

Just a little impact = 32%

No impact at all = 20%

G. "Now I would like to get your reaction to several parts of a possible new immigration bill. Please tell me whether you strongly favor, somewhat favor, somewhat oppose, or strongly oppose each of the following parts of this bill. . . ." "Increasing border security by building a fence along part of the U.S. border with Mexico and by hiring and training more border patrol agents."

Strongly favor = 45%

Somewhat favor = 22%

Somewhat oppose = 13%

Strongly oppose = 20%

H. CBS News/New York Times Poll. June 10-15, 2005. N=1,111 adults nationwide. MoE ± 3 (for all adults). "Do you expect to get back more money than you've contributed to Social Security, less money than you contributed, or about the same amount of money?"

More = 15%

Same = 33%

Less = 52%

I. Gallup Poll. N=1,385 adults nationwide (MoE ± 3), including, with oversamples, 821 non-Hispanic whites (MoE ± 4), 241 blacks (MoE ± 7), and 266 Hispanics (MoE ± 7). Interviewing was June 12-15, 2003, for non-Hispanic whites, and June 12-18, 2003, for blacks and Hispanics.

"Next we have a question about your own experiences as a black. How often do you feel discriminated against in public life or employment because you are black: every day, every week, about once a month, a few times a year, less than once a year, or never?" Options rotated

Every day = 19%

Every week = 8%

Once a month = 13%

Few times a year = 26%

Less than once a year = 15%

Never = 19%

J. CNN/Opinion Research Corporation Poll. May 4-6, 2007. N=1,028 adults nationwide. MoE ± 3.
					.
"Which of the following arrangements between gay or lesbian couples do you think should be recognized as legally valid: same-sex marriages, civil unions but not same-sex marriages, or neither same-sex marriages nor civil unions (outlaw both)?"

Same sex marriage = 26%

Civil unions = 29%

Neither, outlaw both = 45%

2. (12 points) In the Mississippi Poll since 1992, Mississippians have been asked to rate their perceptions of the ideologies of state and national political figures on a five point scale. The scores were 1 for very liberal, 2 for somewhat liberal, 3 for moderate, 4 for somewhat conservative, and 5 for very conservative. What is the mean in words for each of the following persons?

A. Democratic presidential candidate Bill Clinton in 1992 poll, mean = 2.61

B. Democratic presidential candidate John Kerry in 2004 poll, mean = 2.27

C. Republican presidential candidate Bob Dole in 1996, mean = 3.59

D. Republican presidential candidate George Bush in 2000, mean = 3.34

E. Democratic governor Ronnie Musgrove in 2000, mean = 2.71

F. Republican governor Kirk Fordice in 1992, mean = 3.67

3. (8 points) Answer the following questions based on the information about means provided in question 2:

A. Who was perceived to be the most LIBERAL of the six political figures?

B. Who was perceived to be the most CONSERVATIVE of the six political figures?

C. Assume that the average Mississippian’s self-identification mean was a 3.38 in each year. In 1992, was the average Mississippian ideologically closer to Governor Fordice or presidential candidate Clinton?

D. Assume that the average Mississippian’s self-identification mean was a 3.38 in each year. In 2000, was the average Mississippian ideologically closer to Governor Musgrove or presidential candidate Bush?

4A. (5 points) The following two questions are based on the last three Mississippi Polls, all conducted in the 21^st century. Using the statistic of variance, are Democrats or Republicans most divided on each of the following five variables:

A. Affirmative Action:

Democrats’ variance is .813

Republicans’ variance is .544

B. Improving blacks’ socio-economic conditions:

Democrats’ variance is .480

Republicans’ variance is .625

C. Death penalty for murder:

Democrats’ variance is .330

Republicans’ variance is .184

D. Liberal-conservative ideology:

Democrats’ variance is 1.340

Republicans’ variance is 1.018

E. Rate President Bush’ job performance:

Democrats’ variance is .897

Republicans’ variance is .715

4B. (5 points) Using the statistic of variance, are whites or blacks most united on each of the following five variables:

A. Death penalty for murder:

Whites’ variance is .227

Blacks’ variance is .297

B. Liberal-conservative ideology:

Whites’ variance is 1.077

Blacks’ variance is 1.565

C. Income:

Whites’ variance is 5.325

Blacks’ variance is 3.445

D. Party Identification:

Whites’ variance is .207

Blacks’ variance is .106

E. Support for government providing jobs and a good living standard for all:

Whites’ variance is .776

Blacks’ variance is .441

5. (10 points) For each of the following bivariate tables, the predictor is statistically significant at the .05 level. For each of these tables, what is the direction of the relationship (use complete sentences that refer to the categories of the two variables in the form of hypotheses)? These tables examine how age is related to five political attitudes, perceptions, or behaviors in the 2000, 2002, 2004, and 2006 Mississippi Polls.

Table A

Age Differences in Preferences toward State Health Care Spending

AGE

Health Care Spending	18-35	36-55	56 and Over
Less or Same	22%	24%	33%
More	78%	76%	67%
N Size	(794)	(809)	(629)

Gamma = -.174

Chi-squared significance < .001

Note: Cell entries total 100% down each column.

Source: 2000, 2002, 2004, and 2006 Mississippi Polls.

Table B

Age Differences in Seat Belt Usage

AGE

Seat Belt Usage	18-35	36-55	56 and Over
Always	52%	68%	73%
Most of Time	24%	24%	18%
Seldom	17%	5%	6%
Never	7%	3%	3%
N Size	(387)	(361)	(325)

Gamma = -.302

Chi-squared significance < .001

Note: Cell entries total 100% down each column.

Source: 2000, 2002, 2004, and 2006 Mississippi Polls.

Table C

Age Differences in Access to a Personal Computer

AGE

Access to a PC	18-35	36-55	56 and Over
Yes	74%	73%	50%
No	26%	27%	50%
N Size	(597)	(601)	(484)

Gamma = .327

Chi-squared significance < .001

Note: Cell entries total 100% down each column.

Source: 2000, 2002, 2004, and 2006 Mississippi Polls.

Table D

Age Differences in Likelihood of Remaining in Current Community in Five Years

AGE

Likelihood of Staying	18-35	36-55	56 and Over
Definitely No	18%	4%	4%
Probably No	22%	11%	5%
Probably Yes	32%	33%	24%
Definitely Yes	28%	52%	67%
N Size	(384)	(359)	(316)

Gamma = .479

Chi-squared significance < .001

Note: Cell entries total 100% down each column.

Source: 2000, 2002, 2004, and 2006 Mississippi Polls.

Table E

Age Differences in Political Knowledge (able to remember congressman’s name)

AGE

Political Knowledge	18-35	36-55	56 and Over
Low	83%	70%	61%
High	17%	30%	39%
N Size	(798)	(826)	(667)

Gamma = .340

Chi-squared significance < .001

Note: Cell entries total 100% down each column.

Source: 2000, 2002, 2004, and 2006 Mississippi Polls.

6. (5 points) The 2006 Mississippi Poll asked respondents how they would vote in a hypothetical presidential election between Democrat Hillary Clinton and Republican Condi Rice. The gamma coefficients showing the strengths of the relationships between five demographic and attitudinal predictors and the vote follow:

Ideological self-identification gamma = .474

Party identification gamma = .787

Family Income gamma = .515

Race gamma = -.898

Sex gamma = -.124

Regarding the magnitude of these relationships, which predictor is most important in affecting the presidential vote? Which predictor is second in importance? Which is third? Fourth? What is least important?

7. (10 points) A researcher decides to study whether racial differences exist in Mississippi in access to technology (having access to a personal computer). She believes that family income may be an intervening variable that accounts for racial differences in access to technology. Answer the following questions based on the bivariate and multivariate tables below and using the 2000, 2004, and 2006 Mississippi Poll data (some percentages are adjusted to better illustrate the model), which reflect the following model:

FAMILY INCOME

RACE ACCESS TO PC

A. Are there more low income (under $20,000) or high income (over $40,000) people in the sample?

B. In the sample, do you think that more people have access to a personal computer, or lack access to a personal computer?

C. What is the direction of the bivariate relationship between race and access to a personal computer? Use complete sentences that refer to the categories of the two variables in the form of an hypothesis.

D. What is the direction of the bivariate relationship between race and income? Use complete sentences that refer to the categories of the two variables in the form of an hypothesis.

E. What is the direction of the bivariate relationship between income and access to personal computers? Use complete sentences that refer to the categories of the two variables in the form of an hypothesis.

F. Examining low income families alone, is there any statistically significant relationship between race and access to personal computers? If so, what is the direction of the relationship (using complete sentences that refer to the categories of the two variables in the form of an hypothesis)? A magnitude of at least a 5% difference is required for there to be a statistically significant relationship.

G. Examining high income families alone, is there any statistically significant relationship between race and access to personal computers? If so, what is the direction of the relationship (using complete sentences that refer to the categories of the two variables in the form of an hypothesis)? A magnitude of at least a 5% difference is required for there to be a statistically significant relationship.

H. Looking at whites alone, is there any statistically significant relationship between income and access to personal computers? If so, what is the direction of the relationship (using complete sentences that refer to the categories of the two variables in the form of an hypothesis)? A magnitude of at least a 5% difference is required for there to be a statistically significant relationship.

I. Looking at blacks alone, is there any statistically significant relationship between income and access to personal computers? If so, what is the direction of the relationship (using complete sentences that refer to the categories of the two variables in the form of an hypothesis)? A magnitude of at least a 5% difference is required for there to be a statistically significant relationship.

J. Which variable is more important in explaining access to personal computers—income or race?

BIVARIATE: Race .......> Access to personal computers

	White	Black
Have Access to PCs	73%	57%
Don’t Have Access to PCs	27%	43%
	100% (1103)	100% (548)

BIVARIATE: Race .......> Income

	White	Black
Low Income (under $20,000)	20%	46%
Middle Income	31%	34%
High Income (over $40,000)	49%	20%
	100% (1167)	100% (552)

BIVARIATE: Income .......> Access to PCs

	Low Income (under $20,000)	Middle Income	High Income (over $40,000)
Have Access to PCs	38%	69%	88%
Don’t Have Access to PCs	62%	31%	12%
	100% (375)	100% (410)	100% (529)

MULTIVARIATE (Among Low Income Families Only)

	White	Black
Have Access to PCs	40%	36%
Don’t Have Access to PCs	60%	64%
	100% (165)	100% (192)

MULTIVARIATE (Among High Income Families Only)

	White	Black
Have Access to PCs	85%	82%
Don’t Have Access to PCs	15%	18%
	100% (427)	100% (93)

8. (10 Points) A researcher for the state health department is interested in studying what groups are most supportive of establishing universal health care coverage in Mississippi. She uses a question on the Mississippi Poll asking respondents whether they think establishing a universal health care program is a very important priority (coded 4), is important (3), is somewhat important (2), or is not an important priority (coded 1). For each of the following pairs of groups, indicate which group views a universal health care program as the highest priority, and whether the difference between the two groups is statistically significant at the .05 level. (Some data have been artificially manipulated to illustrate important points.) Use means and t-tests.

A. Men have a mean of 2.95. Women have a mean of 3.20. The t-test is statistically significant at the .09 level.

B. Whites have a mean of 2.96. Blacks have a mean of 3.35. The t-test is statistically significant at the .001 level.

C. Liberals have a mean of 3.33. Conservatives have a mean of 2.98. The t-test is statistically significant at the .01 level.

D. Under $10,000 family income people have a mean of 3.40. Over $70,000 family income people have a mean of 2.52. The t-test is statistically significant at the .001 level.

E. Democratic party identifiers have a mean of 3.40. Republican party identifiers have a mean of 2.79. The t-test is statistically significant at the .001 level.

9. (15 points) Circle the MOST correct response regarding the assigned readings.

According to the textbook, what are the differences between GIS and Exploratory Data Analysis?

GIS involves “becoming thoroughly familiar with a set of data” and employing “trimmed means,” while exploratory data analysis “records data that can be displayed geographically.”
Exploratory data analysis involves “becoming thoroughly familiar with a set of data” and employing “trimmed means,” while GIS “records data that can be displayed geographically.”
GIS and exploratory data analysis both involve “becoming thoroughly familiar with a set of data” and employing “trimmed means.”
GIS and exploratory data analysis both involve recording “data that can be displayed geographically.”
Neither GIS nor exploratory data analysis involve “becoming thoroughly familiar with a set of data” by employing “trimmed means,” or recording “data that can be displayed geographically.”

According to the textbook, what is the level of measurement required to use the statistics gamma and the t-test for difference between means?

Gamma requires nominal level and t-test requires ordinal data
Gamma and t-test both require nominal level data
Gamma requires ordinal data and t-test requires interval data.
Gamma requires interval data and t-test requires ordinal data
Gamma and t-test both require interval data

What is the major conclusion of the Norris and Moon article on local government adoption of e-government, entitled “Advancing E-Government at the Grassroots: Tortoise or Hare?”

E-government adoption is progressing rapidly as measured by deployment of Web sites as well as by use of integrated and transactional e-government.
E-government adoption is progressing rapidly if measured solely by deployment of Web sites, but it is progressing much more slowly if measured by movement toward integrated and transactional e-government.
E-government adoption is progressing rapidly as measured by movement toward integrated and transactional e-government, but is progressing much more slowly if measured solely by deployment of Web sites.
E-government is progressing slowly when measured by both deployment of Web sites as well as by use of integrated and transactional e-government.

4. The article on “the Graying of America and Support for Funding the Nation’s Schools” by Eric Plutzer and Michael Berkman reassessed conventional wisdom by conducting a cohort analysis of support for education spending as people aged. What did they find?

Each cohort became more supportive of education spending as they aged
Each cohort became less supportive of education spending as they aged
Each cohort remained unchanged in their attitudes toward education spending
Cohorts born before the Great Depression became less supportive of education spending as they aged, while those born after the Great Depression became more supportive of education spending
Cohorts born before the Great Depression became more supportive of education spending as they aged, while those born after the Great Depression became less supportive of education spending

5. The article by Eva Berman on “Dealing with Cynical Citizens” examined what level of government and with what geographic area?

A. The national government was examined with a national survey

B. State governments were examined with a national survey

C. Local governments were examined with a national survey

D. Attitudes toward the national government was examined with a study of people living in the 11 states of the South

E. Attitudes toward their state governments were examined with a study of people living in the 11 states of the South

(First Example)

RESEARCH METHODS, TEST 2 (Fall 2003)

A. How often do you trust public officials to do what is right?

Almost always = 7%

Most of the time = 31%

Some of the time = 47%

Rarely = 15%

B. Rate the job performance of the state legislature:

Excellent = 7%

Good = 32%

Fair = 49%

Poor = 12%

C. Rate the quality of life in Mississippi:

Excellent = 30%

Good = 50%

Fair = 18%

Poor = 2%

D. Likelihood of living in your community five years from now:

Definitely not = 6%

Probably not = 12%

Probably yes = 34%

Definitely yes = 48%

E. Personal financial satisfaction:

Pretty well satisfied = 41%

More or less satisfied = 40%

Not satisfied at all = 19%

F. Compared to what state and local government in Mississippi is currently spending, how much would you like to spend on environmental programs?

More = 46%

Same = 44%

Less = 10%

G. What is the educational level of the adult population?

High school dropout = 25%

High school graduate = 27%

Some college = 32%

College graduate = 16%

H. What are your own political beliefs?

Very liberal = 6%

Somewhat liberal = 12%

Moderate = 36%

Somewhat conservative = 30%

Very conservative = 16%

I. African-American attitude toward penalty for first-degree murder:

Death penalty = 25%

Life without parole = 69%

Jail term shorter than life = 6%

J. White attitude toward penalty for first-degree murder:

Death penalty = 70%

Life without parole = 29%

Jail term shorter than life = 1%

2. (10 points) Mississippians were asked to rate their perceptions of the ideologies of political figures on a five point scale. The scores were 1 for very liberal, 2 for somewhat liberal, 3 for moderate, 4 for somewhat conservative, and 5 for very conservative. What is the mean in words for each of the following persons?

A. Democratic presidential candidate Al Gore, mean = 2.35

B. Republican presidential candidate George W. Bush, mean = 3.40

C. Senator Trent Lott, mean = 3.84

D. President Bill Clinton, mean = 2.6

E. President George H.W. Bush, mean = 3.5

3. (5 points) Assume that the average Mississippian has an ideological self-identification mean of 3.30. Answer the following questions based on the information in question 2:

A. Is the average Mississippian ideologically closer to Al Gore or to George W. Bush?

B. Is the average Mississippian ideologically closer to President Bill Clinton or to President George H.W. Bush?

C. Which of the five political figures mentioned in question 2 is perceived by Mississippians to be most liberal?

D. Which of the five political figures mentioned in question 2 is perceived by Mississippians to be most conservative?

E. Which of the five political figures mentioned in question 2 is perceived by Mississippians to be closest to their own ideological position?

4. (10 points) What is the range for each of the following indicators?

A. A feeling thermometer that ranges from a low score (cold) of 0 to a high score (hot) of 100.

B. An ideological self-identification scale that ranges from a 1 for strong liberal to a 5 for strong conservative.

C. An age indicator that ranges from a low of 18 to a high of 99.

D. A years lived in Mississippi indicator that ranges from a low of 1 to a high of 99.

E. An agree-disagree item pertaining to abortion that ranges from 1 for strong agreement to 4 for strong disagreement.

F. A family income indicator that ranges from 10,000 to 100,000.

G. A state spending indicator that is coded as: 1 for more, 2 for same, and 3 for less.

H. A party identification indicator that ranges from 1 for strong Democrat to 7 for strong Republican.

I. A job performance indicator that is coded as: 1 is excellent, 2 is good, 3 is fair, and 4 is poor.

J. A tax increase for education item that is coded as 1 for supporting the tax increase and 2 for opposing the tax increase.

5. (5 points) Using the statistic of variance, are whites or blacks most divided on each of the following five variables:

A. Number of adults in household:

White variance is .601

Black variance is 1.655

B. Age of respondent:

White variance is 291

Black variance is 286

C. Education level:

White variance is 1.228

Black variance is 1.203

D. Family income level:

White variance is 5.191

Black variance is 3.187

E. Ideological self-identification:

White variance is 1.114

Black variance is 1.554

6. (5 points) Using the statistic of variance, are women or men most united or homogeneous on each of the following five variables:

A. Number of adults in household:

Men variance is .910

Women variance is 1.061

B. Age of respondent:

Men variance is 349

Women variance is 273

C. Education level:

Men variance is 1.309

Women variance is 1.357

D. Family income level:

Men variance is 5.192

Women variance is 4.987

E. Ideological self-identification:

Men variance is 1.209

Women variance is 1.169

7. (10 points) For each of the following bivariate tables, indicate whether the predictor is statistically significant (.05 level), and if so, what the direction of the relationship is (use complete sentences that refer to the categories of the two variables in the form of hypotheses)?

Table A

Age Differences in Attitude towards a Woman’s Role in Society

AGE

WOMAN’S ROLE	18-35	36-55	56 and Over
Not Equal	12%	16%	30%
Equal	82%	84%	70%
N Size	(618)	(708)	(466)

Gamma = -.359

Chi-squared significance < .01

Note: Cell entries total 100% down each column.

Source: 1998, 1999, 2000 Mississippi Poll.

Table B

Income Differences in Attitude towards a Woman’s Role in Society

FAMILY INCOME

WOMAN’S ROLE	< $20,000	$20-40,000	$40-60,000	> $60,000
Not Equal	25%	15%	18%	8%
Equal	75%	85%	82%	92%
N Size	(413)	(480)	(280)	(224)

Gamma = .261

Chi-squared significance < .001

Note: Cell entries total 100% down each column.

Source: 1998, 1999, 2000 Mississippi Poll.

Table C

Ideological Differences in Attitude towards a Woman’s Role in Society

SELF-IDENTIFIED IDEOLOGY

WOMAN’S ROLE	Liberal	Moderate	Conservative
Not Equal	12%	15%	22%
Equal	88%	85%	78%
N Size	(382)	(596)	(726)

Gamma = -.243

Chi-squared significance < .001

Note: Cell entries total 100% down each column.

Source: 1998, 1999, 2000 Mississippi Poll.

Table D

Race Differences in Attitude towards a Woman’s Role in Society

RACE

WOMAN’S ROLE	White	African-American
Not Equal	17%	21%
Equal	83%	79%
N Size	(1193)	(554)

Gamma = -.150

Chi-squared significance < .05

Note: Cell entries total 100% down each column.

Source: 1998, 1999, 2000 Mississippi Poll.

Table E

Sex Differences in Attitude towards a Woman’s Role in Society

SEX

WOMAN’S ROLE	Men	Women
Not Equal	18%	17%
Equal	82%	83%
N Size	5%	5%

Gamma = .016

Chi-squared significance < .893

Note: Cell entries total 100% down each column.

Source: 1998, 1999, 2000 Mississippi Poll.

8. (5 points) Regarding the magnitude of the relationships in question 7, which predictor is most important in affecting attitude towards a woman’s role in society? Which predictor is second in importance? Which is third? Fourth? What is least important? Use the gamma statistic.

9. (10 points) Two researchers decide to study what factors influenced voting in the last presidential election. They decided to focus on party identification and people’s attitudes toward the impeachment of President Clinton. Answer the following questions based on the bivariate and multivariate tables below:

A. Are there more Democrats in party identification or more Republicans in the sample?

B. In the sample, do more people support the impeachment of Clinton or oppose the impeachment?

C. What is the direction of the bivariate relationship between party identification and the presidential vote?

D. What is the direction of the bivariate relationship between party identification and one’s attitude towards the impeachment of President Clinton?

E. What is the direction of the bivariate relationship between one’s attitude towards the impeachment of Clinton and the presidential vote?

F. Examing Democrats alone, is there any relationship between attitude towards the impeachment of Clinton and the presidential vote? If so, what is the direction of the relationship?

G. Examing Republicans alone, is there any relationship between attitude towards the impeachment of Clinton and the presidential vote? If so, what is the direction of the relationship?

H. Looking at those who supported impeachment, is there any relationship between party identification and the presidential vote? If so, what is the direction of the relationship?

I. Looking at those who opposed impeachment, is there any relationship between party identification and the presidential vote? If so, what is the direction of the relationship?

J. Which variable is more important in explaining the presidential vote—party identification or attitude towards the impeachment of Clinton?

BIVARIATE:Party Id .......> Presidential Vote

	Democratic Party Id	Republican Party Id
Gore (Dem) Vote	80% (400)	10% (30)
Bush (Rep) Vote	20% (100)	90% (270)
	100% (500)	100% (300)

BIVARIATE: Party Id .......> Attitude toward Impeachment of Clinton

	Democratic Party Id	Republican Party Id
For Impeachment	10% (50)	83% (250)
Against Impeachment	90% (450)	17% (50)
	100% (500)	100% (300)

BIVARIATE: Attitude toward Impeachment .......> Presidential Vote

	For Clinton Impeachment	Against Impeachment
Gore (Dem) Vote	22% (65)	73% (365)
Bush (Rep) Vote	78% (235)	27% (135)
	100% (300)	100% (500)

MULTIVARIATE (Among Democrats Only)

	For Impeachment	Against Impeachment
Gore (Dem) Vote	80% (40)	80% (360)
Bush (Rep) Vote	20% (10)	20% (90)
	100% (50)	100% (450)

MULTIVARIATE (Among Republicans Only)

	For Impeachment	Against Impeachment
Gore (Dem) Vote	10% (25)	10% (5)
Bush (Rep) Vote	90% (225)	90% (45)
	100% (250)	100% (50)

10. (10 Points) A researcher for the College Board is interested in studying what groups are most supportive of spending more money of public higher education in Mississippi. Attitude toward higher education is coded as follows: 1= spend more; 2 = spend the same as now; 3 = spend less than now. For each of the following pairs of groups, indicate which groups prefers to spend more on higher education, and whether the difference between the two groups is statistically significant at the .05 level. Use means and t-tests.

A. Men have a mean of 1.45. Women have a mean of 1.37. The t-test is statistically significant at the .009 level.

B. Whites have a mean of 1.48. Blacks have a mean of 1.25. The t-test is statistically significant at the .001 level.

C. Liberals have a mean of 1.34. Conservatives have a mean of 1.49. The t-test is statistically significant at the .001 level.

D. High school dropouts have a mean of 1.32. College educated have a mean of 1.48. The t-test is statistically significant at the .001 level.

E. Young adults under 35 have a mean of 1.41. Older adults over 55 have a mean of 1.42. The t-test is statistically significant at the .653 level.

11. (10 points) Circle the correct response regarding each of the assigned articles.

1. Which of the following best describes the findings for: In Welfare and Work: Job-Retention Outcomes of Federal Welfare-to-Work Employees, by Gooden and Bailey?

A. Welfare-to-Work employees have less of a chance of retaining their jobs than non-Welfare-to-Work employees.

B. Welfare-to-Work employees have the same chance of retaining their jobs as non-Welfare-to-Work employees.

C. Welfare-to-Work employees have a greater chance of retaining their jobs than non-Welfare-to-Work employees.

D. None of the above

2. In the article In Welfare and Work: Job-Retention Outcomes of Federal Welfare-to-Work Employees, by Gooden and Bailey, a majority of employees in the Welfare-to-Work category tend to be of which demographic categories?

A. Men and blacks

B. Women and blacks

C. Men and whites

D. Women and whites

3. In the article Mainframe and PC Computing in American Cities: Myths and Realities by Norris and Kraemer, in comparing cities with central computer systems and those using solely personal computers, which of the following findings was made?

A. Central system cities were more widely automated

B. Central system cities had more widespread use among staff

C. Central system cities were more likely to deploy leading edge computer technologies

D. All of the above

E. None of the above

4. In the article Dealing with Cynical Citizens by Evan Berman, which measure of statistical significance is used most frequently?

A. Cramer’s V

B. t-test

C. Pearson r significant test

D. Regression coefficient significance level

E. None of the above

5. What were the data sources of the articles Mainframe and PC Computing in American Cities: Myths and Realities by Norris and Kraemer and Dealing with Cynical Citizens by Evan Berman?

A. Both employed national surveys with cities or city leaders

B. Norris and Kraemer used a national survey and Berman studied cities only in the South

C. Norris and Kraemer studied only cities in the South, while Berman used a national survey

D. Both examined cities or city leaders only in the South

(Second Example)

POLITICAL ANALYSIS, TEST 2, SPRING 2004 (91 point test)

1. (10 points) What is the mode category for each of the following distributions of Mississippians on questions asked in the 2002 Mississippi Poll? Circle the correct category in each of the following ten examples.

A. Support for government help to improve social and economic positions of blacks and other minorities:

1) Strongly agree- 22%

2) Agree- 53%

3) Disagree- 19%

4) Strongly disagree - 6%

B. Government should cut spending, even if education and health programs must be cut:

1) Strongly agree- 6%

2) Agree- 33%

3) Disagree- 49%

4) Strongly disagree - 12%

C. Government should guarantee jobs and a good standard of living to everyone:

1) Strongly agree- 16%

2) Agree- 46%

3) Disagree- 30%

4) Strongly disagree - 8%

D. What is your ideological self-identification?

1) Very liberal- 9%

2) Somewhat liberal- 14%

3) Moderate- 31%

4) Somewhat conservative- 29%

5) Very conservative- 17%

E. Last year, what was your total family income?

1) Under $10,000- 11%

2) $10-20,000- 17%

3) $20-30,000- 18%

4) $30-40,000- 16%

5) $40-50,000- 9%

6) $50-60,000- 8%

7) $60-70,000- 6%

8) Over $70,000- 15%

F. Party Identification:

1) Strong Democrat- 22%

2) Weak Democrat- 14%

3) Independent Democrat- 9%

4) Pure Independent- 8%

5) Independent Republican- 12%

6) Weak Republican- 16%

7) Strong Republican- 19%

G. Place of Residency Size:

1) Farm- 9%

2) Rural area- 31%

3) Under 2,500- 12%

4) 2,500-10,000- 18%

5) 10-50,000- 19%

6) Over 50,000- 11%

H. Rate the quality of life in Mississippi:

1) Excellent- 26%

2) Good- 48%

3) Fair- 20%

4) Poor- 6%

I. How much you trust Mississippi public officials to do what is right:

1) Always- 6%

2) Most of time- 35%

3) Some of time- 47%

4) Never- 12%

J. Should abortion be outlawed?

1) Strongly agree- 27%

2) Agree- 31%

3) Disagree- 28%

4) Strongly disagree- 14%

2. (10 points) What is the median category for each of the ten examples in question 1? Write the number of the correct category next to each of the letters of the question below:

3. (10 points) In 2000 the Mississippi Poll surveyed voters’ ideological orientations, and their perceptions of the presidential candidates’ and Senator Trent Lott’s ideologies. The ideology and ideological perception variables were coded as follows:

1- Very Liberal

2- Somewhat Liberal

3- Moderate

4- Somewhat Conservative

5- Very Conservative

The mean of voters’ own ideologies was a 3.26

The mean of voters’ perceptions of Al Gore’s ideology was 2.35

The mean of voters’ perceptions of George Bush’s ideology was 3.40

The mean of voters’ perceptions of Trent Lott’s ideology was 3.84

From this information on means, answer the following five questions. Use one complete sentence for each question. That sentence should be in the following format: “Voters own ideology/or their perception of a candidate’s ideology was somewhere between _____ and _____, but closer to ______.” The blanks are the names of ideology categories listed above.

A. What is the ideology of the average voter?

B. What is the average voter’s perception of Al Gore’s ideology?

C. What is the average voter’s perception of George Bush’s ideology?

D. What is the average voter’s perception of Trent Lott’s ideology?

E. Which presidential candidate was closer in perceived ideology to the average voter? One word answer. Was it Al Gore, or George Bush?

4. (10 points) What is the range of each of the following variables? Give a one number response for each of the following variables.

A. The salaries at a public university in the South range from $10,000 per year for a janitor to $200,000 for a university President.

B. The number of times poll respondents have been arrested for alleged DUI offenses ranges from zero to five.

C. Party identification is coded from a low of 1 for Strong Democrats to a numerically high code of 7 for Strong Republicans.

D. How warmly or coldly do you feel towards President Bush? Rate him anywhere on a feeling thermometer from a low of 0 if you feel very cold toward him to a high of 100 if you feel very warm towards him. Give him intermediate scores depending on how warm or cold you feel towards him.

E. Age of respondents in the 2002 Mississippi Poll. The youngest is 18, and the oldest is 92.

5. (10 points) This is a question about variance. There is much debate regarding which race is more united or divided in opinions on political issues—whites or African Americans. For each of the following variables for which the variance of blacks and whites is listed, which of the two races is more united in attitudes. Give a one word answer for each variable- white, or black.

A. Party Identification.

White variance = 4.2

Black variance = 2.5

B. Ideological self-identification.

White variance = 1.1

Black variance = 1.6

C. Affirmative action opinions.

White variance = 0.5

Black variance = 0.7

D. Abortion opinions.

White variance = 2.1

Black variance = 2.0

E. Gubernatorial general election vote.

White variance = 0.3

Black variance = 0.1

6. (10 points) What is the direction of the relationship between sex and a political opinion for each of the following bivariate tables? Give a one sentence response for each table, and use the language of an hypothesis. Assume that all of the relationships are statistically significant.

IDEOLOGY	MEN	WOMEN
Liberal	18%	27%
Moderate	32%	28%
Conservative	50%	45%
(N Sizes)	(265)	(293)

Gamma = -.14

PARTY IDENTIFICATION	MEN	WOMEN
Democrats	28%	44%
Independents	34%	25%
Republicans	38%	31%
(N Sizes)	(265)	(293)

Gamma = -.23

EDUCATION LEVEL	MEN	WOMEN
Less than high school	22%	25%
High School Graduate	60%	58%
College Graduates	18%	17%
(N Sizes)	(289)	(316)

Gamma = -.05

ABORTION	MEN	WOMEN
Outlaw It	61%	54%
Keep it Legal	39%	46%
(N Sizes)	(279)	(301)

Gamma = .16

LIKELY VOTERS	MEN	WOMEN
Non Voters	36%	40%
Voters	64%	60%
(N Sizes)	(290)	(318)

Gamma = -.08

7. (10 points) List the five variables above that are related to sex in order of the magnitude of their relationships with sex. List the most important variable first, and the least important variable fifth, and include the other three variables in the proper order of their importance. Overall, is sex strongly related to these five politically relevant variables, or is it not strongly related, and explain your answer in one sentence?

8. (21 points) This is a two part question, which is listed on the next page.

	MEN	WOMEN
ALWAYS USE SEAT BELTS	43%	59%
SELDOM USE SEAT BELTS	57%	41%
(N SIZES)	(565)	(634)

Relationship is statistically significant at .05 level.

	HIGH SCHOOL DROPOUTS	HIGH SCHOOL GRADUATES	COLLEGE GRADUATES
ALWAYS USE SEAT BELTS	43%	51%	68%
SELDOM USE SEAT BELTS	57%	49%	32%
(N SIZES)	(381)	(616)	(200)

Relationship is statistically significant at the .05 level.

	MEN	WOMEN
HIGH SCHOOL DROPOUTS	32%	33%
HIGH SCHOOL GRADS	54%	56%
COLLEGE GRADUATES	14%	11%
(N SIZES)	(4291)	(4889)

Relationship is not significant at the .05 level.

TABLE OF HIGH SCHOOL DROPOUTS ONLY

	MEN	WOMEN
ALWAYS USE SEAT BELTS	32%	51%
SELDOM USE SEAT BELTS	68%	49%
(N SIZES)	(175)	(199)

Relationship is statistically significant at the .05 level.

TABLE OF HIGH SCHOOL GRADUATES ONLY

	MEN	WOMEN
ALWAYS USE SEAT BELTS	41%	59%
SELDOM USE SEAT BELTS	59%	41%
(N SIZES)	(279)	(337)

Relationship is statistically significant at the .05 level.

TABLE OF COLLEGE GRADUATES ONLY

	MEN	WOMEN
ALWAYS USE SEAT BELTS	62%	74%
SELDOM USE SEAT BELTS	38%	26%
(N SIZES)	(104)	(95)

Relationship is statistically significant at the .05 level.

A major public policy problem is the extent to which people do not use their seat belts, and are therefore more likely to be severely injured in automobile accidents. Answer the following questions with a complete sentence for each question (except F and G, where you circle the correct responses). Remember that if a relationship is not statistically significant, then it does not really exist. These data are from the Mississippi Poll over a twenty year period.

A. Is there a relationship between sex and seat belt usage, and if so what is the direction of the relationship (hypothesis format)? Rely on a bivariate table.

B. Is there a relationship between education level and seat belt usage, and if so what is the direction of the relationship? Rely on a bivariate table.

C. Is there a relationship between sex and education level, and if so what is the direction of the relationship? Rely on a bivariate table.

D. Turning to the three multivariate tables, is there a relationship between sex and seat belt usage after controlling for education level? If so, what is the direction of the relationship between sex and seat belt usage for each level of education?

E. Comparing across the three multivariate tables and assuming that the relationships are statistically significant, what is the direction of the relationship between education level and seat belt usage among men? What is the direction of the relationship between education level and seat belt usage among women?

F. Based on your analysis of the multivariate tables, which of the two variables (if either) has a direct effect on seat belt usage? Is the answer sex, education, both, or neither (circle the correct answer)?

G. Which of the following four groups is most likely to wear their seat belts? (circle the correct group) Is it? Women who are high school dropouts; women who are college graduates; men who are high school dropouts; or men who are college graduates.

(Third Example)

POLITICAL ANALYSIS, TEST 2, SPRING 2002 (75 point test)

1. (10 points) The Mississippi Poll asked a number of issues of whites and blacks during the late 1990s. For each of the following examples, what is the mode?

A. Among whites, 23% want to permit abortion, 69% want to outlaw it, and 8% are undecided.

B. Among blacks, 18% want to permit abortion, 79% want to outlaw it, and 3% are undecided.

C. Among whites, 79% want government to provide cheap health care, and 21% are against governmental action in this area.

D. Among blacks, 87% want government to provide cheap health care, and 13% are against governmental action in this area.

E. Among whites, 46% support government making every effort to improve blacks’ social-economic positions, and 54% are against government helping blacks.

F. Among blacks, 92% support government making every effort to improve blacks’ social-economic positions, and 8% are against government helping blacks.

G. Among whites, 11% are for affirmative action, and 89% are against it.

H. Among blacks, 60% are for affirmative action, and 40% are against it.

I. Among whites, 70% support the death penalty for first degree murderr, and 30% are for life in prison without parole.

J. Among blacks, 25% support the death penalty for first degree murder, 69% favor life in prison without parole, and 6% support a prison term shorter than life.

2. (10 points) What is the median category for each of the following examples, drawn from the Mississippi Poll in the late 1990s?

A. Age:

18-29 = 23%

30-45 = 36%

46-59 = 20%

60 and over = 21%

B. Education:

High school dropout = 25%

High school graduate = 27%

Some college = 32%

College graduate = 16%

C. Income:

Under $20,000 = 30%

$20,000-$40,000 = 34%

$40,000-$60,000 = 20%

Over $60,000 = 16%

D. Party Identification:

Democrats = 47%

Independents = 11%

Republicans = 42%

E. Trust in State Public Officials?

Almost Always = 7%

Most of the Time = 31%

Some of the Time = 47%

Rarely = 15%

3. (10 points) In the year 2000 the Mississippi Poll asked voters what their own ideology was, and what they perceived the ideologies of three politicians to be. The numbers assigned to the ideology variable were:

Very Liberal = 1

Somewhat Liberal = 2

Moderate = 3

Somewhat Conservative = 4

Very Conservative = 5

In words, what is the Mean score for each of these four variables?

A. Voters’ Own Ideological Views, mean = 3.26

B. Voters Perceived Al Gore’s Ideology, mean = 2.35

C. Voters Perceived George W. Bush’s Ideology, mean = 3.40

D. Voters Perceived Trent Lott’s Ideology, mean = 3.84

E. Was the average voter’s ideology closer to their perception of Al Gore’s, or closer to George W. Bush’s?

4. (10 points) What is the range for each of the following variables?

A. The age at inauguration of American Presidents ranged from a low of 42 for Theodore Roosevelt to a high of 73 for Ronald Reagan in his second term.

B. The total number of staff members in the President’s Executive Office ranged from a low of 683 in 1944 under FDR to a high of 5,721 in 1972 under Nixon.

C. From 1953 through 1992, the President’s success rate in dealing with Congress has ranged from a low of 52% under George Bush Senior to a high of 85% under John Kennedy.

D. Presidential vetoes cast have ranged from a low of 0 under five Presidents in the 19^th century to a high of 635 under Franklin D. Roosevelt.

E. The number of Supreme Court nominations that failed in the Senate has ranged from a low of 0 under a number of Presidents to a high of 5 under Tyler.

5. (10 points) Our 2001 NSF Grassroots Party Activists study examined the ideologies of Democratic and Republican party members in Mississippi. The ideology scores were:

1 = Very Liberal

2 = Somewhat Liberal

3 = Moderate

4 = Somewhat Conservative

5 = Very Conservative

We determined the mean and variance scores of both parties, and of whites and blacks as follows:

GROUP MEANS VARIANCE

Democrats 2.69 1.36

Republicans 4.45 0.49

Whites 4.03 1.11

Blacks 2.37 1.41

A. Which of the two political parties– Democrats or Republicans– is most divided and diverse in terms of its ideology? Which political parties was more conservative?

B. Which of the two races– whites or blacks– is most divided and diverse in terms of its ideology? Which race was more conservative?

C. Which of these four groups– Democrats, Republicans, whites, or blacks– tends to be the most liberal?

6. (15 points) For each of the following bivariate relations, answer three questions: is there is a statistically significant relationship between the two variables, what is the magnitude of that relation, and what is the direction of the relation (in hypothesis format)?

A. Age and Education

	Age 18-35	Age 36-55	Over 55
High School Dropout	24%	16%	36%
High School Graduate	34%	35%	29%
Some College	30%	25%	20%
College Graduate	12%	24%	15%

Chi Squared = 0

Gamma = -.06

B. Age and Attitude Toward Government Creating Public Jobs

	Age 18-35	Age 36-55	Over 55
Strongly Favor	31%	21%	21%
Somewhat Favor	46%	45%	41%
Somewhat Oppose	21%	28%	31%
Strongly Oppose	2%	6%	7%

Chi Squared = .01

Gamma = .197

C. Age and Ideology

	Age 18-35	Age 36-55	Over 55
Liberal	32%	24%	14%
Moderate	32%	35%	32%
Conservative	36%	41%	54%

Chi Squared = .02

Gamma = .18

D. Age and Federal Medical Care

	Age 18-35	Age 36-55	Over 55
Strongly Favor	41%	32%	26%
Favor	49%	53%	60%
Oppose	10%	15%	14%

Chi Squared = .05

Gamma = .19

E. Age and Sexist?

	Age 18-35	Age 36-55	Over 55
Sexist	2%	4%	10%
Not Sexist	98%	96%	90%

Chi Squared = 0

Gamma = -.50

7. (10 points) A hypothetical study at MSU finds that women faculty have lower mean salaries than male faculty. The AAUW, a women's rights group, files a class action lawsuit charging that male administrators are discriminating against women because of their gender. The university's lawyer contends that women faculty are on average paid less than men because of their lower seniority--that most women faculty have worked for the university less time than men. Examine the following series of multivariate tables, which examine the relations between sex and salary for three different groups of workers based on their years of seniority. Which of these two arguments, if either, appears to be supported by the data? Explain and defend your answer fully.

TABLE CONTAINING THOSE WITH LESS THAN FIVE YEARS OF SERVICE

	Men	Women
Under $40,000 Salary	80%	80%
Over $40,000 Salary	20%	20%
Totals	(100)	(250)

TABLE CONTAINING THOSE WITH FIVE TO TWELVE YEARS OF SERVICE

	Men	Women
Under $40,000 Salary	55%	55%
Over $40,000 Salary	45%	45%
Totals	(100)	(100)

TABLE CONTAINING THOSE WITH OVER TWELVE YEARS OF SERVICE

	Men	Women
Under $40,000 Salary	10%	10%
Over $40,000 Salary	90%	90%
Totals	(100)	(50)