Sociology 364/497Name: ______
Spring 2011
FINAL EXAMINATION
Section I - The What of Science
Part I - Multiple Choice (one point each):
1. Science assumes all but which of the following.
A. There exists a reality that can be measured.
B. Knowledge is more important than ignorance.
C. Variables are causally related
D. Metaphysical concepts can be measured
2. "Operational Definitions" can vary from researcher to researcher.
A. True
B. False
3. If a person receives an "RFP" to study an aspect of terrorism, and pursues the research, he is likely influenced by what motivation in doing so:
A. Personal
B. Intellectual
C. Institutional
D. None of the above
4. Interval level measurement uses numbers while nominal level measurement uses numerals.
A. True
B. False
5. A researcher indexes the rise in unemployment among students by comparing the proportional change in the number of students employed now versus last year. This is an example of the use of what descriptive statistic:
A. Frequency
B. Percentage
C. Ratio
D. Rate
6. A coefficient of equivalence is likely to be higher than a coefficient of concurrent.
A. True
B. False
7. The best measure of central tendency to use when the a distribution is skewed is the:
A. Mean
B. Median
C. Mode
D. Standard Deviation
Part II - Definitions (Two points each)
1. Indicator:
2. Three-Quarter High Rule:
3. Standard Score:
Part III - Matching (one point each)
1. ______Mean for Grouped Dataa.
2. ______Standard Deviation for Grouped Datab.
3. ______Median for Grouped Datac.
Section II - The How of Science
Part I - Multiple Choice (one point each)
1. Snowball samples are better to use in predictive research than are cluster samples
A. True
B. False
2. Which of the following is not a probability sample:
A. Simple
B. Systematic
C. Referral
D. Stratified
3. Alpha is the probability of retaining the Null Hypothesis when it is false.
A. True
B. False
4. An experimental design lacking an equivalent control group is likely what type design:
A. Simple
B. Complex
C. Quasi
D. None of the above
5. Complex experimental designs allow researchers to study "interaction effects"
A. True
B. False
6. A 3x3x3 CRD indicates all but which of the following:
A. Three independent variables
B. Eighteen cells
C. Nine conditions
D. Three interaction effects
Part II - Definitions (two points each)
1. Disproportionate Stratified Random Sampling
2. Central Limit Theorem
3. Latin Square Design
Part III - Matching (one point each)
1. ______Z-ratioa.
2. ______t-ratiob.
3. ______ANOVAc.
4. ______Confidence Intervald.
Section III - The Where of Science
Part I - Multiple Choice (one point each)
1. “How many miles do you commute from home to CSUN” is what type of survey question:
A. Information
B. Opinion
C. Demographic
D. Self Perception
2. One reason for using structured questions is that they are easier to code:
A. True
B. False
3. According to our discussions in class, which form of data collection yields the lowest
response rate:
A. Field Interviews
B. Telephone Interviews
C. Mail Out Questionnaires
D. They Are All The Same
4. The difference between a Complete Observer and Complete Participant is that the former is more covert than the latter..
A. True
B. False
5. Which of the following is the best example of a Prisoner's Dilemma
A. Class Grades
B. The Stock Market
C. A Marriage
D. A Jury
6. Phi is an appropriate measure of association for two interval level variables:
A. True
B. False
Part II - Definitions (two points each)
1. Demographic Question
2. Sociometry
3. Thomas
Part III - Matching (one point each)
1. ______Phia.
2. ______Standard Errorb.
3. ______Pearson's rc.
4. ______Regressiond.
Section IV - The When of Science
Part I - Multiple Choice (one point each)
1. If the partial crosstab tables show values for the measure of association that are lower than
the original zero order table, we would likely say there is an antecedent relationship.
A. True
B. False
2. Which of the following is least likely to happen in a partial correlation:
A. The partial correlation = the zero order correlation
B. The partial correlation is less than the zero-order correlation
C. The partial correlation is more than the zero-order correlation
D. They are all equally likely
3. The expected frequency in a cross-classification table is obtained by the formula (RxC)/T.
A. True
B. False
4. Which of the following is the best synonym for a "Beta Weight".
A. Path Coefficient
B. Standardized Partial Regression Coefficient
C. Partial Correlation Coefficient
D. All of the above are synonyms
5. The Introduction of a research report contains the "hypotheses"
A. True
B. False
6. Which of the following sections would most likely contain a listing of the references for a
research paper:
A. Abstract
B. Results
C. Appendix
D. Discussion
Part II - Definitions (two points each)
1. Suppresser variable:
2. Partial Correlation:
3. Beta Weight:
Part III - Matching (one point each)
1. ______Indirect Effectsa.
2. ______R2b.
3. ______Direct Effectsc.
4. ______Partial rd.
Short Answer (Four points each) Answer only five of the following six.
1. List the four approaches to social theory discussed in class and state the data gathering
method each would use.
2. Describe the steps for turning a concept into a variable.
3. Discuss the three types of variation found in the F-ratios of a complex experimental design.
4. Name the four possible outcomes in assessing partial tables and state what those outcomes mean.
5. Describe the steps in performing a path analysis?
6. Articulate the procedure of recoding a variable from a Likert answer format (i.e. SA to SD)
into a dichotomy, in SPSS.
Problems (four points each). Answer only four of the following five problems
1. Find the mean and standard deviation for the following set of data.
i | f
2 | 4
3 | 7
4 | 9
5 | 5
6 | 5
2. You want to study the effects of lower division statistics prerequisites taken, on subsequent grades in sociology 364, so you compile information about the grades of students who have had one of three exposures to lower division statistics categories Test the null hypothesis that the statistic prerequisite taken does not affect subsequent grade in Sociology 364 (use alpha < .01, Fcrit = 5.90).
None Taken Another College Math 140 @ CSUN
647280
Grades in Soc 364687684
728088
768492
808896
3. Given the following sample data from 200 students showing the crossclassification of GPA by gender at CSUN, calculate Phi and test the hypothesis that these data are independent in the population (alpha = .05, =3.841)
.
Hi GPA Lo GPA
------
Male| 30 |70 |100
------
Female| 70 |30 |100
------
Total 100 100200
4. The Law School Aptitude Test (LSAT) is used to predict subsequent performance in LawSchool. Assume the correlation between the LSAT and subsequent GPA is 0.5. If a student scored 700 of the LSAT, what would be his predicted GPA, if the mean LSAT is 500 with a standard deivation of 100 and the mean GPA for graduate students is 3.2 with a standard deviation of 0.2. How accurate would be your prediction (Hint: remember the formula for the standard error of the estimate).
5. You know Grades and intelligence are correlated and, indeed, find that rxy = 0.6 between GPA (Y) and IQ (X). But you sense it has more to do with motivation and tenacity than native brain capacity. So, you collect data to find that the correlation between GPA (Y) and time spent studying (Z) to be ryz = 0.8 and the correlation between IQ (X) and time spent studying (Z) to be rxz = 0.75. What can be said about your intuition?
Bonus Question #1 - (3 points): What is the definition of Description Statistics discussed in class?
Bonus Question #2 - (3 points): Name and give the difinition for six of the twelve vocabulary words mentioned in class.
1.
2.
3.
4.
5.
6.
Bonus Question #3 - (3 points): Identify the personal statement made the first day of class by each of these six students:
1. Kimberly -
2. Brandon -
3. Rocio -
4. Leticia -
5. Marcus -
6. Robert -