Research Methods in Social Psychology

1
Research Methods in Social Psychology
2
Last Time

• psychology relies on empirical methods
• theories are general propositions about causal
  relationships amongst constructs
• hypotheses are conceptual statements that require
  operationalization
• science is a process of conjecture and
  refutation

3
Our Proposition

• Hypothesis Southern white males are more prone
  to aggression than are Northern white males.
  (conjecture)
• Well look at different strategies to examine
  hypotheses. (refutation)

4
Refutation

• putting theories to the test
• What evidence supports (and, more importantly,
  disconfirms) our theories?

5
Descriptive Research

• assesses the amount or average level of a given
  variable in a population
• e.g., public opinion surveys
• not a true test of an hypothesis
• What is the white male homicide rate in the South?

6
Descriptive Research

• What is the white male homicide rate in the South?

adapted from Nisbett (1993)
7
Descriptive Research

• critical issues
• random sampling
• basis of comparison
• informative, but not a hypothesis test
• What else could account for the findings?

8
Correlational Research

• investigates whether changes in one variable are
  related to changes in another variable
• What is the relationship between being from the
  South and aggressive behavior?

9
Correlational Research

• correlation coefficients
• range from 1.00 to -1.00
• positive correlation increase/decrease in the
  same direction

10
Correlational Research
11
Correlational Research

• correlation coefficients
• range from 1.00 to -1.00
• negative correlation increase/decrease in
  opposite directions

12
Correlational Research
13
Correlational Research

• correlation coefficients
• range from 1.00 to -1.00
• strength of the relationship closeness to
  1.00/-1.00, not by the valence (/-)
• Which indicates a stronger correlation -.74 or
  .21?

14
Correlational Research
15
Correlational Research

• What is the relationship between being from the
  South and aggressive behavior?
• Nisbett (1993)
• homicide rate and southernness rs .37 .52
• in the social sciences (Cohen, 1992)
• r .50 (strong)
• r .30 (moderate)
• r .10 (small)

16
Correlational Research

• strengths of this approach
• can be a random sample
• actual behavior
• has good generalizability (i.e., external
  validity)
• potential for numerous variables

17
Correlational Research

• weakness of this approach
• cannot infer a cause-effect relationship
• factors for inferring causality
• association -- yes
• temporal priority -- no
• rule out a spurious relationship -- no

18
Correlational Research

• temporal priority

time
southernness
homicide rates
19
Correlational Research

• temporal priority

time
southernness
homicide rates
or
time
homicide rates
southernness
20
Correlational Research

• temporal priority
• direction of causality problem

time
southernness
homicide rates
or
time
homicide rates
southernness
21
Correlational Research

• rule out a spurious relationship

another variable
southernness
homicide rates
spurious
22
Correlational Research

• rule out a spurious relationship

poverty
rs .38 .42
southernness
homicide rates
(Nisbett, 1993)
23
Correlational Research

• rule out a spurious relationship
• 3rd variable problem

another variable
southernness
homicide rates
spurious
24
Correlational Research

• strengths
• potential for numerous variables
• good generalizability
• weaknesses
• cannot make causal conclusions
• direction of causality
• 3rd variable problem

25
Addressing Problems with Correlational Research

• direction of causality problem
• Does southernness lead to more aggression or
  does a propensity for aggression lead to more
  southernness?
• solution we cause (i.e., manipulate) one of the
  variables
• insult one group on Southerners, but not another

26
Addressing Problems with Correlational Research

• direction of causality problem
• if we control who is insulted, then measure
  aggressiveness, we know the direction of
  causality
• but, we still have the 3rd variable problem
• SES
• poor social skills

27
Addressing Problems with Correlational Research

• random assignment to condition
• if SES or social skills have an effect on
  aggression, it should be equal for both groups
• We can address the problems of correlational
  research by doing experiments.

28
Experimental Design

• characteristics
• manipulation of a variable
• solves the direction of causality problem
• randomly assign to conditions
• solves the 3rd variable problem

29
Experimental Design

• language of experiments
• independent variable
• dependent variable
• operational variable
• random assignment

30
Experimental Design

• testing theory
• Does the independent variable cause changes in
  the dependent variable?

cause
southernness
aggression
31
Experimental Design

• the Southern culture of honor hypothesis
• Cohen et al. (1996)
• 2 (Southern/Northern) X 2 (insult/no insult)
• chicken, cortisol, testosterone

32
Experimental Design
Distance at which participants gave way to
confederate (inches)
Insult
No Insult
33
Experimental Design
Distance at which participants gave way to
confederate (inches)
Insult
No Insult
34
Experimental Design
Percentage Change in Cortisol Level
Insult
No Insult
35
Experimental Design
Percentage Change in Cortisol Level
Insult
No Insult
36
Experimental Design
Percentage Change in Testosterone Level
Insult
No Insult
37
Experimental Design
Percentage Change in Testosterone Level
Insult
No Insult
38
Experimental Design

• strengths
• allows for causal conclusions to be made -- best
  test of theory
• weaknesses
• not all questions are amenable to experiments
• concerns about generalizability

39
Methodological Challenges

• expectancy effects
• demand characteristics of the situation
• social desirability concerns
• ethical dilemmas

40
Summary

• refutation process of testing theories
• descriptive research is informative, but limited
  in theory testing
• correlational research is more informative, but
  does not allow for causal explanations
• experiments are the best test of theories

41
Next Time

• social cognition
• how we think about the social world