Quasi Experimental and single case experimental designs

1
Quasi Experimental and single case experimental
designs

• Chapter 11

2
Program Evaluation

• Research on programs that are proposed and
  implemented to achieve some positive effect on a
  group of people
• i.e. DARE program evaluation
• Applies research approaches to evaluate types of
  programs

3
Program Evaluation

• Types
• Needs assessment
• What problems, if any, need to be addressed?
• Once identified researchers can plan accordingly
• Assessment of program theory
• Used to fix problems found in a needs assessment
• Involves collaboration of service providers,
  researchers, clientele, etc
• Process evaluation
• Program monitoring to determine if program is
  doing what it is supposed to
• Is the target population being reached?
• Are the research questions being answered?

4
Program Evaluation

• Types
• Outcome evaluation
• Impact assessment
• What did the participants like/dislike about the
  study?
• Studying the outcome and then assess the impact
  of the outcome measure.
• Efficiency assessment
• Was the program worth it?
• Did the benefits outweigh the costs?
• Uses quasi-experimental designs to assess
  effectiveness and efficiency

5
Quasi experimental designs

• 1 group posttest only design
• Has no comparison group
• Doesnt have good internal validity
• 1 group pretest-posttest design
• History
• Could confound results
• Maturation
• People change over time and this could skew
  results
• Testing
• Previous experience could change behavior
• Instrument decay
• the deterioration of research instruments or
  methods during a study

6
Quasi experimental designs

• Instrument decay
• equipment may wear out, respondents may become
  more casual in recording their responses.
• Regression toward the mean
• a.k.a statistical regression
• The tendency for extreme scores in a distribution
  to move (regress) toward the mean of the
  distribution w/ repeated testing.
• Problem of reliability
• Occurs when a set of extreme scores collected at
  one time point is compared to scores taken at
  another time point.

7
Quasi experimental designs

• Nonequivalent control group design
• A separate control group
• Selection differences choosing groups based on
  naturally existing categories this is why study
  is called nonequivalent
• Nonequivalent control group pretest-posttest
  design
• Benefit a pretest for comparison purposes
• Interrupted time series design
• Measurement of dependent variable is interrupted
  by the quasi independent variable
• By taking multiple measurements you can look at
  the effectiveness of a treatment before and after
  it is introduced.
• Control series design
• Extension of the interrupted time series design
  with a control group

8
Quasi experimental designs

• 1 group posttest only design
• E X O
• 1 group pretest-posttest design
• E O X O
• Nonequivalent control group design
• E X O
• C O
• Nonequivalent control group pretest-posttest
  design
• E O X O
• C O O
• Interrupted time series design
• E O O O X O O O
• Control series design
• E O O O X O O O
• C O O O O O O

E experimental group X treatment C control
group O observation to collect data
9
Single case experimental designs

• A.K.A. Single subject designs or Single
  participant designs
• Baseline control period ? treatment period
• The change in behavior during these periods could
  be attributed to a variety of things
• Reversal designs
• To determine if treatment had an effect, revert
  back to baseline (no treatment) and look at the
  differences
• ABA design
• Baseline control period ? treatment period ?
  control period
• Can also do a ABAB design

10
Single case experimental designs

• Multiple baseline designs
• Observing behavior before and after a
  manipulation under multiple circumstances
• Can be done
• Across subjects
• Across behaviors
• Across situations