TEST I -- 1997 APLS 301 Dr

TEST I APLS 301 Dr. Botsch NAME ________________________________

Note: You may earn up to 5 BONUS POINTS for neatness on this test (which already has 103 points on it).

Honor Pledge: I have neither given nor received any unauthorized aid on this examination.

_______________________________________________

signature

I. Evaluate the hypotheses listed below. For each one, describe the following in the spaces provided. (20 points)

a) the unit of analysis

b) the independent variable

c) the dependent variable

d) flaws, errors, or problems in the hypothesis (in other words, does the hypothesis meet the criteria for a good hypothesis, and if not, then what is wrong with it?)

1. In comparing open-seat elections for the U.S. House of Representatives, the candidate with the most money tends to win.

2. Among Aiken County Republican voters, self-identification as a southerner is positively associated with their position on abortion.

3. In comparing lottery games, the size of the payoff increases profits more than advertising or administrative efficiency.

4. The proposed war on Iraq will hurt the nation more than help the nation.

5. Among USCA students, regardless of gender, age is positively associated with levels of knowledge about American government.

II. We talked about a number of possible sources of knowledge that we all use in everyday life as well as in science at the beginning of the course. What is the source of knowledge of each of the following statements (sensory, intuition, authority, or logic)? Discuss the likelihood that the statement or argument is true or factually correct (e.g. the statement may be unlikely to be true because the logical form is incorrect, or one has no way of knowing because it is a value judgment). (10 points)

1. The Constitution gives Congress the sole power to declare war.

2. According to my political science professor, the real hard part of a unilateral war with Iraq will be after we win the military battle, not the battle itself.

3. I think that the stock market will rebound just in time to help Republicans win control of both houses of Congress. (Note: the "I" in this statement refers to you, the student, not to your professor!)

4. Premise 1: All students hate tests.

Premise 2: This is a test.

Premise 3: You are a student.

‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑‑

Conclusion: Therefore, you hate this test.

5. Coding the gender of the respondent in the telephone interview as female from the sound of the voice.

III. Short answer questions. (14 points)

1. Stating what is already known is called the _______________________________

stage of (or step in) scientific research.

2. Laying out the relationship you expect to find is the ___________________________ stage of (or step in) scientific research.

3. Rethinking what was already known after the research is completed is called the __________________________ stage or step in scientific research.

4. Research is thought to be objective and value free. Where (in what step) do values inevitably enter into the steps of the scientific research process? Why?

5.Why must we have a theory BEFORE we test hypotheses? (Hint: what would happen if we just started testing hypotheses using all possible relationships?)

6. Explain why social scientists say they “reject the null hypothesis” rather than accept the hypothesis that they really want to test.

7. Why is replication important? (I will be looking for 2 reasons.)

IV. Calculating sample error. (8 points)

1. Suppose you do a telephone survey calling 500 of the 60,000 people registered to vote in Aiken County 2 months before the November 5, 2002 election. You select households to call using a systematic sample of the phone book and then adding 1to each number selected and then use the “last birthday” method to choose the person to interview in each household. Your response rate was 60%. You found that 65% favor Sanford for governor. What is the sampling error in this finding? Explain your findings in terms of the confidence interval and significance level. (Show all your work.)

2. Suppose we interviewed 500 customers at the local WalMart one Wednesday afternoon to find out what people in Aiken like about shopping at WalMart. We use systematic sampling to choose them as they leave the store. Assume that there are well over 50,000 people who regularly shop at this WalMart. What will the sampling error be for this survey? (Show all your work.)

V. Suppose you are asked to design survey questions for the Aiken SmartGrowth interest group to find out whether or not residents of our fair city wish to have more restrictions on land development in order to preserve green space and reduce congestion. Of course, the cost of these restrictions would be higher development costs and some loss in the freedom of individual landowners to do with their property what they wish. You volunteered for this job because you successfully completed the notorious research methods class taught by the highly respected political scientist Bob Botsch. The club president has jotted down some questions for you to work on. Using the guidelines for writing reliable and valid survey questions, evaluate the questions and answer sets she gave you, as shown below. If there is nothing wrong, say it is "ok." If there are problems, describe what they are and rewrite the question to improve it. Pay attention to layout as well as question and answer content! (8 points)

1. Do you think that our local government should do more to protect green space and reduce congestion? ___ 1) yes ___ 2) no ___ 3) don't know

2. There is not such thing as a free lunch. Anything local government does to protect green space will necessarily adversely impinge upon those who currently own land that they may wish to develop. In thinking about tradeoffs here, which is more important to you, the rights of landowners to develop their land as they see fit or the rights of the community at large and future generations to preserve green space that enhances all property values over the long run (Circle your answer.)

1) property rights 2) preserving green space 3) have not had time to think about this

3. How do you feel about the following statement? The noxious fumes of traffic congestion and the incredible waste of human lives as people sit in traffic is a crime against humanity and the livability of our fair city. (Circle your response.):

1) strongly agree 2) agree 3) have mixed feelings 4) disagree 5) strongly disagree

6) have not had time to think about

4. How often do you get caught in traffic jams while driving in Aiken? (Circle the number next to your response.)

1) almost daily 2) rarely 3) several times a week 4) don’t know

VI. Suppose you were hired by United Way America (the national organization) to design the sample selection procedure for a survey of social service agency heads (like those directing mental health centers, emergency hotlines, teen pregnancy prevention centers, family Y’s, etc.) all over the nation that are funded by local United Way agencies. The goal is to find out how they rate their local United Way agency in terms of being fair and objective in making funding decisions. You need a sample of 500 and have no one to help you on the survey except one local college student who is working as a volunteer. You have a budget of $2,500 (not counting your own salary). You also have an electronic list of names and addresses and phone numbers of all 10,000 agencies and their heads funded by United Way across the nation. (8 points)

1. What kind of survey (face to face, telephone, or mail) should you do and why?

2. Explain how you would draw your sample, that is how you would choose each respondent to interview and why.

3. Explain how you would try to maximize the reliability of this survey? (You should make several points here, but only give ones relevant to the type of survey you are doing!))

4. How would you try to maximize the response rate? (Again, several!)

VII. Weighted samples. (8 points)

Suppose the city government in which you work asks you to do a survey for them. They are interested in changing the hours at which the local library is open. They want to see if there is sufficient support to change the existing hours. You currently have a list of all households that receive water bills (which includes every household in the city – 60,000 households). You also have a computerized list of existing library patrons – all 8,000 of them.

1. If you want to be able to compare the opinions of each of the general population who are not library patrons (you will screen out patrons in doing telephone interviews in the general population subsample) with those who are already patrons with about a + or ‑ 4% error for each subgroup, how large a subsample would you need for each group? (To keep this simple, ignore the correction factor in making this calculation. Show your calculations.)

2. Suppose you did the survey using equally weighted subsamples (of the size you just calculated) that approximated simple random samples, and found that 55% of the general population want the change, but only 40% of the patrons support the change. Draw a conclusion about whether the groups really differ in supporting the change.

3. Based on the numbers you used in part 1, show how you can reweight the two subsamples to produce an accurate combined sample for the whole population. Produce such a projection and draw a conclusion. Show all your work.

VIII. Levels of Measurement. (4 points) Indicate the level of measurement of each of the following (Nominal, Ordinal, Interval, Ratio, or Dichotomous)

1. Per pupil state funding of higher education in dollars: __________________

2. Type of zoning laws in place: mixed use zoning, non-mixed use zoning, and each of these types with and without impact fees attached to them: _________________________

3. Ratings of states in access to higher education: excellent, good, fair, or poor: __________________

4. If a course is traditional in format or nontraditional: ____________________

IX. Concepts, Definitions, and operational measures. (4 points)

1. Suppose we are concerned with the concept of “self-identification as a religious fundamentalist.” How appropriate is the following definition? Self-identified religious fundamentalists are those who believe that the Bible should be interpreted literally and is the completely accurate word of God. Explain your answer.

2. Suppose we decide to measure self-identified religious fundamentalism in our exit poll by asking the following question: Do you believe that the Bible is the word of God? If they say “yes,” they are coded as self-identified religious fundamentalists and if they say “no,” they are identified as non-fundamentalists. Evaluate this operational measure in terms of both its reliability and its validity.

X. Evaluate this test as a compound measure in terms of the following areas. Briefly explain your answers. (12 points)

1. reliability

2. validity

3. independence

4. homogeneity

5. coverage

6. discrimination

XI. Assign the proper codes (use 1 through 5 as codes) to the answers of the following items so that they can be combined into a Likert scale that measures degree of “nerdness." Give the highest codes to the most nerdy oriented responses. Using the answers that are indicated by x's, compute and record the score for this one individual. (5 points)

1. I like to wear a plastic pocket protector in my shirts.

___ strongly agree

_x_ agree

___ undecided

___ disagree

___ strongly disagree

2. Football is fund to watch.

___ strongly agree

___ agree

___ undecided

_x_ disagree

___ strongly disagree

3. I always read the book before I see the movie.

_x_ strongly agree

___ agree

___ undecided

___ disagree

___ strongly disagree

4. I love to do my APLS301 homework.

___ strongly agree

___ agree

_x_ undecided

___ disagree

___ strongly disagree

The scale score using the simple counting or adding method for this person would be _______. (Show your computations and explain why you did it the way you did.)

Suppose this respondent failed to answer the third question. How could we score her so that we didn't have to throw out the interview? Show what her score would be and show your work.

XII. Explain why the four item measure above is better than using the fourth item or question alone. (2 points)