Transcript A Sociological Compass

Sociology:
Your Compass for a New World
Robert J. Brym and John Lie
Wadsworth Group/Thomson
Learning
© 2003
Chapter 2
Research Methods
Science and Objectivity
The Aim Of Science Is To Arrive At
Knowledge That Is Less Subjective Than
Other Ways Of Knowing.
Objectivity Is Achieved By Testing Ideas
Against Systematically Collected Data And
Leaving Research Open To Public Scrutiny.
Science and Subjectivity
Creativity And The Motivation To Study
New Problems From New Perspectives
Arise From Individual Passions And
Interests.
How Research Filters Perception
Values
Theories
Previous Research
Methods
“Reality”
The Research Cycle
Figure 2.2
The Research Cycle
1. formulate question
6. report results
2. review existing
literature
5. analyze data
3. select method
4. collect data
Methodological Issues
Reliability:
Consistency In Measurement
Validity:
Precision In Measurement
Generalizability:
Assessing The Relevance Of Findings Beyond
The Case Studied
Causality:
Assessing Cause-And-Effect Relations Among
Measurement as Target Practice
1.Not Valid, Not Reliable
x
x
x
x
2. Not Valid, Reliable
3. Valid, Not Reliable
xx
xx
4. Valid, Reliable
xx
xx
x
x
x
x
5. Valid, Reliable, Generalizable (Target 2)
xx
xx
Variables
Variables Are Concepts That Can Assume
More Than One Value.
In a Cause-and-effect Relationship
The Independent Variable Is the Cause
The Dependent Variable Is the Effect
The Influence of the Control Variable Is
Removed From the Relationship Between
the Independent and Dependent Variables
Participant Observation
Involves Observing People’s Face-to-face
Interactions and Participating in Their Lives
Is Useful for Exploratory Research,
Constructing Grounded Theory, and Validating
Measures on the Basis of Internal Criteria
Issues of External Validity, Reliability,
Generalizability, and Causality Make It Less
Useful for Other Purposes
Experiments
Isolate Hypothesized Causes and Measure
Their Effects by
Randomizing the Allocation of Subjects to
Experimental and Control Groups and
Exposing Only the Experimental Group to
an Independent Variable.
They Rank High on Reliability and the
Analysis of Causality but in Sociology They
Have Validity and Generalizability Problems.
Surveys
Ask People Questions About Their Knowledge,
Attitudes or Behavior
In a Face-to-face or Telephone Interview
In a Paper-and-pencil Format.
They Rank High on Reliability and Validity.
Generalizability Is Achieved by Probability
Sampling, Statistical Control, and the Analysis of
Causality by Data Manipulation.
Turning a Classroom
into a Contingency Table
BACK
10 or fewer hours TV
per week and no act of
physical violence per
year
more than 10 hours TV
per week and no act of
physical violence per
year
LEFT
RIGHT
10 or fewer hours TV
per week and at least 1
act of physical violence
per year
more than 10 hours TV
per week and at least 1
act of physical violence
per year
FRONT
Testing an Association
for Spuriousness
(1) We believe there is a causal relationship between TV viewing & aggressiveness:
TV viewing
(independent variable)
(association)
Aggressiveness
(dependent variable)
(2) By controlling for gender we can see whether gender has created a spurious
association between TV viewing and aggressiveness:
TV viewing
(independent variable)
Respondent’s gender
(control variable)
(no association)
aggressiveness
(dependent variable)
Existing Documents
and Official Statistics
Inexpensive and Convenient Sources of
High-quality Data Must Be Used
Cautiously
They Often Reflect the Biases of the
Individuals and Organizations That Create
Them Rather Than the Interests of the
Researcher
Correlation
Dependent variable
Dependent variable
Dependent variable
60
60
60
r = .85
40
r=0
r = -.92
40
40
20
20
20
0
0
0
2
4
6
Independent Variable
Positive Correlation
8
0
0
2
4
6
8
Independent Variable
Negative Correlation
0
2
4
6
Independent Variable
No Correlation
8