Matching (2 points each)
|
Terms |
Letter of Matching Definition |
|
1. General Social Survey |
I |
|
2. Statistical Abstract of the United States |
B |
|
3. Nonreactive Research |
A |
|
4. Conceptualization |
G |
|
5. Operationalization |
C |
|
6. Likert Scale |
F |
|
7. Non-Probability Sampling |
H |
|
8. Probability Sampling |
D |
|
9. Inferential Statistics |
J |
|
10. SOC 300 |
E |
Definitions
a. Research in which the participants do not know
they are being studied. The data are such things as family portraits, old
letters, trash, tombstones, and public behavior.
b. An annual publication of government collected
data.
c. The process of moving from
a definition of an abstract idea to a way to observe or measure it empirically.
d. When researchers randomly select eligible people
or cases for a study in a way in which all have a chance of being chosen.
e. A class that is now over!
f. A scale often used in survey research in which
people express attitudes or other responses in terms of several ordinal-level
categories (strongly agree to strongly disagree).
g. The process of developing clear, rigorous,
systematic definitions of abstract ideas.
h. When researchers select people or cases for a
study in a way in which all do not have a chance of being chosen.
i. An annual
or biannual survey of a random sample of approximately 1500 U.S. adults.
j. A type of quantitative analysis in which
researchers analyze sample data and generalize to a population.
Fill-in-the-Blank (3 points each): Identify the level of
measurement of each of the items below.
|
11. Education as: less than High School, High
School Degree, Some College, College Degree, Graduate Classes or Degree ORDINAL |
|
|
12. Marital Status NOMINAL |
16. Race
NOMINAL |
|
13. Age in Years INTERVAL |
17. Do you smoke: Yes or No NOMINAL or ORDINAL |
|
14. Number of Siblings INTERVAL or RATIO |
18. Fear of Crime Frequency: a great deal, somewhat, not often,
never ORDINAL |
|
15. Health Status:
Excellent, Good, Fair, Poor ORDINAL |
19. U.S. Region of residence: East, West, North,
South NOMINAL |
20. Professor Bob studied class level and drinking behavior
at Wild State University in 1999. He used the 1999 student telephone directory,
and after a random start selected every 15th student. He then mailed a
questionnaire to the 1,000 students selected and had two follow-up postcard
reminders. What kind of sampling is being used?
a.
Cluster sampling
b.
Systematic random sampling
c.
Simple random sampling
d.
Purposive sampling
21. In the question above, what is the sampling frame
of the study?
a. All freshmen.
b. All students.
c. All students in the telephone book.
d. All students who live on campus.
22. Dr. Horse
developed a measure of an ideal place to live. He added together measures of
many factors: tax rate, quality of school system, cultural and recreational opportunities,
pollution, traffic congestion, crime rate and health care availability for 100
U.S. cities to get a score for each. Dr. Horse created a(n):
a. Index
b. Scale
c. Measure of central tendency
d. Item analysis
23. Dr. Lui
wants to measure fear of crime. He develops a question for a qualitative
interview that asks, “Have you ever been the victim of a crime?” What is one of the major problems with this
measure of fear of crime?
a.
Reliability
b.
Validity
c.
Stability
d.
Triangulation
24. While developing
a survey, you decide to ask two questions on people’s attitudes about
socialized day care. You will compare individual answers to the two questions
to see if they are similar. What issue in measurement are you trying to assess:
a. Reliability
b. Validity
c.
Stability
d. Triangulation
25. Which of
the following is FALSE about secondary data analysis?
a. A gap may exist between a researcher's
conceptualization of a variable and how it is measured in available data.
b. Locating data with specific variables of
interest can be time consuming and sometimes a researcher may not make data
available.
c. Information about how data was collected may
be insufficient to determine whether there is bias.
d. It is very expensive compared to equivalent
primary data collection.
26. Claude
DuPere has a list of measures on the French influence in the New Orleans area.
He asked you to identify the one that is NOT an unobtrusive measure.
Which one is it?
a. A survey using a three-page questionnaire partly
written in French that was distributed to residents of a neighborhood.
b. Walking down a street in New Orleans and
noticing that most of the signs in stores in a neighborhood are in French or
French-Cajun.
c. A list of votes supporting bills on bilingual
education in the Louisiana state legislature with the area represented by the
legislator noted on the list.
d. A box of 300 letters written by people living
in New Orleans to relatives living in French speaking areas outside the state
(e.g., Quebec) between 1980 and 1985.
27. What is
the problem with this measure of college student age?
Please indicate your age:
17 to 19 years old
19 to 21 years old
21 to 23 years old
a. Its responses are double-barreled.
b. Its responses are at a nominal level of
measurement.
c. Its responses are not mutually exclusive.
d. There is nothing wrong with it.
28. Which is a
major ADVANTAGE of content analysis?
a. It is time consuming and requires a large
staff with specialized equipment.
b. It is unobtrusive.
c. It cannot measure how people experience the
"text" or how the text affects them.
d. It cannot be used to study materials, which
are not written or recorded.
29. Which sequence illustrates the progression of
quantitative measurement steps?
a. Conceptualization, operationalization,
empirical measurement.
b. Operationalization, conceptualization,
empirical measurement.
c. Empirical measurement, conceptualization,
operationalization.
d. Empirical measurement, operationalization,
conceptualization.
30. This test fairly reflects the course
readings, lectures and discussion on unobtrusive research, measurement, and
sampling. (No wrong answer)
a.
True
b.
False
Essay (20 points): Write an essay answer on
ONE of the following questions, 1 page in length.
·
What are the limitations of using existing data/existing statistics?
The
data is not collected for the purposes of studying your research question,
which can lead to problems in validity and reliability in using the data to
answer your research question. The questions may not be worded as you would
have worded them, or the response sets may not match what you would have used
in order to answer your research question.
For example, if you wanted to study race issues and wanted to be able to
include Asian and Native American Indians in your analysis but the question
only includes “Black”, “White” and “Other” as responses. Or you want to study men’s use of
pornography and the closest question you can find in the dataset is whether
people have purchased a pornographic magazine in the last year (ideally a
measure of pornography consumption would include more than just magazines, such
as movies, Internet sites, etc..)
The
data may not be of the unit of analysis that you want to study. For example,
you want to study why people leave their jobs, but the dataset only includes
measures of the organization such as turnover rate, number of employees, etc..
and no questions asked of individual employees.
The
sample may not represent the groups you would like to generalize too. For
example if you wanted to be able to talk about all Americans, but the sample
underrepresents people who live in rural areas.
The
data may have a lot of missing data, especially if it was collected via
government agencies for government purposes.
·
Why are multiple indicators usually better than one indicator?
Abstract
concepts are often difficult to measure with just one question. For example,
how would you measure happiness, quality of life, or health status with just
one question?
Using
multiple indicators of these concepts allows you to measure dimensions of a
concept, such as someone’s physical and emotional health. Or dimensions of quality of life such as job
satisfaction, relationship satisfaction, and health status.
Using
multiple indicators improves validity and reliability, and allows you to assess
the validity and reliability of the measures.
·
Describe the different types of nonprobability samples and identify when
each is appropriate to use.
Convenience
sample: You pick people for the study
by convenience to the researcher. For
example, asking any people on campus to respond to a survey. You use this sample design when you have
limited time or resources and do not intend to generalize or develop an
explanation from the sample.
Quota
= You pick people for the study by convenience to the researcher, but get an
apriori determined number of men or women, or freshmen, sophomores, juniors and
seniors, etc.. For example, the campus
convenience sample above but with 50% male and 50% female. You use this sample
design when you have limited time or resources and do not intend to generalize
or develop an explanation from the sample, but would like to try to represent
some of the diversity that exists in the population. You pick the quota
characteristics that you think are related to your research question. For example, with the campus survey, if you
think gender influences people’s answers to the questions, then use gender
quotas.
Purposive:
You choose people or organizations who are eligible for your study by going to
where you can find these people or organizations. There is no list of them
available from which to randomly sample.
Or it doesn’t make sense to try to find one or use one. For example, you are study how children play,
so you go to playgrounds and watch them.
You wouldn’t necessarily need to make a list of all the playgrounds and
randomly select playgrounds to observe. You may want to make sure you watch
playgrounds in different socio-economic areas.
You select these playgrounds intentionally. Or
Theoretical:
You choose people or organizations as specified by your theory that you are
trying to test or build. For example,
if you are trying to determine how businesses owned by feminists operate (in
terms of human resources, policies, economic success, etc..), you would find
such businesses and try to study them.
· Are sampling frame problems
avoided using random-digit-dialing? Explain? What is the population in a study
using random-digit-dialing?
The population with random-digit-dialing samples is
all telephone numbers in the exchanges you are sampling. Sampling frame problems are not avoided
entirely. Some problems that are
avoided include not having unlisted numbers or wrong numbers on the sampling
frame (such as a phone book), and the fact that you don’t have to find a list
of numbers to sample off of. Some
problems that remain with RDD sampling is that households without phones are
not included, households with multiple
phones are overrepresented, households with multiple dwellers but only one
phone are underrepresented, non-residential numbers are contacted (and often
are not eligible for the study), and some numbers are no longer in use.