Research Design And Validity

1
Research Design And Validity
2
Reasons to Conduct a Research Study

• Exploratory data collection
• Hypothesis testing
• Causal relationships can be tested
• Correlational relationships can be tested

3
Experimental Research

• 4 defining characteristics
• Manipulation of variable(s) (IV)
• Measurement of the behavior (DV)
• Group/Condition design
• Control over extraneous variables

4
How to Control Extraneous Variables?

• Hold extraneous variables constant
• Randomize their effects across conditions

5
Pros and Cons of Experimental Approach

• Pros ability to identify causal relationships
• Cons cannot use this approach if variables
  cannot be manipulated

6
Correlational Research

• Goal is to determine whether two variables
  co-vary and to establish the direction of the
  relationship
• No variables are manipulated
• Two or more variables measured - predictor
  variable and criterion variable

7
Positive Aspects of Correlational Research

• Gathering data in exploratory stages of research
• Examining situations where you cannot manipulate
  variables
• Useful for prediction

8
Problems with Correlational Research

• Third-variable problem
• Directionality problem

9
Internal Validity

• The ability of the research design to adequately
  test the hypotheses
• Extraneous variables may interfere with internal
  validity

10
How to Establish Internal Validity?

• You could get two groups that are identical and
  administer treatment to one
• But, there are problems in getting identical
  groups
• Selection bias problems
• 3 types of selection bias
• Self-selection bias
• Experimenter selection bias
• Arbitrary rule selection bias

11

• YOU COULD match two groups on every
  characteristic you think is problematic
• But, there may be too many variables to match
• YOU COULD get P's, measure them, make sure
  nothing changes in the meantime, measure them
  again (pretest-posttest method)
• But, there are a number of threats to internal
  validity in this situation

12
7 Threats to Internal Validity (Campbell
Stanley, 1963)

• History
• Maturation
• Testing
• Instrumentation
• Statistical regression
• Biased selection of participants
• Experimental mortality

13
External Validity

• The degree that results generalize beyond the
  experimental situation
• Threats to external validity
• Artificiality (using highly controlled lab
  conditions)
• Reactivity in the testing situation
• Subject selection biases

14
Internal Validity vs. External Validity

• They trade off against each other
• The researcher must determine which is more
  important

15
Research Settings

• Choice of setting affected by
• Type of research question asked
• Convenience
• Cost
• Ethical considerations

16
Laboratory Setting

• Any setting that is artificial relative to the
  setting where the behavior naturally occurs
• Advantage better control over extraneous
  variables
• Disadvantage loss of external validity?
• One way to deal with this simulation
• Simulation - re-create a real world situation in
  the laboratory

17
Stanford Prison Study

• Zimbardo (1971)
• Clip
• web site

18
Simulations and Realism"

• Mundane realism
• Experimental realism

19
Field Setting

• Research conducted in the participant's natural
  environment
• Advantage greater external validity
• Disadvantage less control over extraneous
  variables