QuasiExperimental Designs

1
Quasi-ExperimentalDesigns

• Chapter 11

2
Research in applied settings

• True experiments may not be possible
• Still need to assess whether an independent
  variable had an effect on a dependent variable
• Does 3 strikes reduce crime?
• Does D.A.R.E. work?
• Does the new employee training program reduce
  missed work days?

3
Program Evaluation

• Needs Assessment
• Program Theory Assessment
• Process Evaluation
• Outcome Evaluation
• Efficiency Assessment

4
Nonequivalent control group pretest-posttest
design

• O X O
• -----------------
• O O
• Treatment and control groups may not be
  equivalent no random assignment
• Compare change in the two groups

5
Interrupted Time Series Design

• O O O O O O X O O O O O O
• Change from immediately before to after treatment
  is difficult to interpret
• Allows clearer interpretation of effect of
  treatment

6
Control Series Design

• O O O O O O X O O O O O O
• O O O O O O O O O O O O
• Improves on the times series design
• Observations in control condition reduce
  ambiguities that may be present in the treatment
  observations

7
Single-subject Designs

• N1 Design
• Single-Person Design

8
Reversal Design

• Establish a baseline for behavior (A)
• At a random point, introduce treatment and
  continue observing behavior (B)
• At a random point, withdraw the treatment (A)
• Is there a reversal of the behavior?

9
ABAB Reversal Design

• Improves on ABA design
• A second treatment change in behavior is powerful
  evidence for its effect

10
Multiple Baseline Design

• Do not withdraw treatment
• Show evidence through multiple replications
• Multiple subjects
• Multiple behaviors
• Multiple situations

11
Developmental Research

• Studying Changes Across Age

12
Cross-Sectional

• Study people of different ages at one time
• Ex Compare people 30, 50, 70 years of age
• Problem cohort effects may explain age
  differences

13
Longitudinal

• Study the same people as they grow older
• Ex Find the same people when they are 30, 50,
  and 70
• Examine changes with age
• Expensive and difficult

14
Sequential method

• Combine cross-sectional and longitudinal
  approaches
• Take cross-sectional approach at time 1 but
  follow up the same people for a longitudinal
  study at time 2
• Provides immediate age difference information