Single Subject Design

1
Single Subject Design Statistical Tests

• Psyc 3705
• Thanks to John Stanton for some of the material
  is this PowerPoint!

2
1-1 Be able to tell me the function of a
research design and data analysis

• Research is designed in such a way to
• 1. Allow for the effect of the independent
  variable to be apparent in the dependent variable
• 2. Control for extraneous variables and other
  possible explanations for the change in the DV
• 3. Present/analyze the data in a format that
  makes sense for interpreting what occurred in the
  study.

3
1-2 Be able to describe in general terms what a
single subject research design is.

• Single subject research designs
• Typically examines one participant at a time to
  investigate the effects of an IV on a DV (e.g. a
  treatment on some behavior of that participant.
• Generally will use multiple (i.e. 6-10
  participants) which are basically replications of
  the research each time an additional participant
  is used.

4
Example of SSD
5
1-3. Know the characteristics of SSD.

• Uses repeated measures
• Need reliable measurement/instrumentation
• Requires a clear description of conditions and
  the DV
• Measurement, IV and DV operationally defined
• 2 general types of phases (could have probes,
  etc.)
• Baseline and treatment conditions

6
1-4 Be able to list and describe the 2 phases of
a SSD.

• 1. Baseline
• Period of no treatment (IV) or traditional
  treatment (reflects natural state)
• Allows researched to have a comparison for the
  effect of IV and to determine if extraneous
  variables are operating (i.e. the control
  condition)

7
Phases of SSD (continued)

• 2. Treatment
• Introduction of the IV
• Phase length should be approximately at least as
  long as baseline (for comparison purposes)
• Repeated measurement of the DV continues

8

• JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2008, 41,
  131135 NUMBER 1 (SPRING 2008)
• TRAINING TO INCREASE SAFE TRAY CARRYING AMONG
• COCKTAIL SERVERS
• MEGAN D. SCHERRER AND DAVID A. WILDER
• FLORIDA INSTITUTE OF TECHNOLOGY

9

• JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2007, 40,
  223237 NUMBER 2 (SUMMER 2007)
• TEACHING MATH SKILLS TO AT-RISK STUDENTS USING
• HOME-BASED PEER TUTORING
• KRISTIN H. MAYFIELD AND TIMOTHY R. VOLLMER

10
1-5 Know the general rules when using single
subject designs

• Shouldnt change baseline until
• Little variability
• No trends
• Unless the trend is in the opposite direction of
  expected effect
• Only change one variable at a time
• (as you will not know which one caused any change
  in the DV!)

11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
1-6. Know the General rules for SSD

• Length of phase (suggestions)
• Keep all the similar length
• Controls for maturation
• To reduce carry over effects use shorter phases
  (treatment)
• Extend phases to include cyclic subject variables
• i.e. menstrual cycle, daily issues (weekends
  Monday)

16
Types of SSD

• Reversal design
• Alternating treatments design
• Multiple baseline design
• Changing criterion design

17
1-7 Know the critical features for the reversal
design and be able to create a graph that
illustrates the reversal design that I specify
(i.e. AB, ABA, ABAB)

• .
• At least 2 conditions
• Baseline
• Treatment
• Same condition present for consecutive sessions
• Conditions dont alternate rapidly
• Types
• AB
• ABA
• ABAB

18
1-8. Be able to recognize and provide examples
of an AB design, as well as know the benefits and
problems with such a design.

• 1 baseline and 1 treatment phase
• Ex. Kid having trouble working on school work
  (i.e. staying on task), the treatment is setting
  up a reinforcement contingency that gives him a
  dollar for every minute he stays on task

19
AB
A
B
20
AB

• Problems
• Limited control over threats to internal validity
• No control for extraneous variables - that is
  changes in the DV could be caused by numerous
  things

21
1-9. Be able to recognize and provide examples
of an ABA design, as well as know the benefits
and problems with such a design.

• Baseline (A) Treatment (B) Baseline (A)
  (withdrawal)
• The withdrawal strengthens the relationship
  between the IV and DV
• Ex. Kid having trouble working on school work
  (i.e. staying on task), the treatment is setting
  up a reinforcement contingency that gives him a
  dollar for every minute he stays on task

22
ABA
A
B
A
23
ABA

• Problems
• Doesnt completely control for extraneous
  variables (but more evidence!)
• Irreversibility
• Some IVs cant be withdrawn and some behaviors
  cant be reversed
• Ethical and legal issues
• Sequence effects

24
1-10. Be able to recognize and provide examples
of an ABAB design, as well as know the benefits
and problems with such a design.

• Baseline Treatment Baseline Treatment
• Increases the confidence in the treatment effect
• Better controls for extraneous variables
• Ex. Kid having trouble working on school work
  (i.e. staying on task), the treatment is setting
  up a reinforcement contingency that gives him a
  dollar for every minute he stays on task

25
ABAB
A
B
A
B
26
ABAB

• Problems
• Irreversibility (same as in ABA)
• Sequence effects

27
1-11 Know the critical features of the
alternating treatments design.

• 2 or more treatments presented
• Treatment phases are rapidly alternated
• In a pseudo random fashion
• Generally a baseline phase (not required)

28
Alternating treatments
29
Alternating treatments

• Advantages
• No treatment withdrawal
• Faster comparison of treatments
• Variable data is okay
• Problems
• Artificiality
• Interference between treatment phases
• Treatments must sufficiently differ

30
1-12 Know the critical features of the multiple
baseline design.

• Critical features
• More than one baseline
• recorded during the same time period for
  different lengths of time
• Introduce treatment in one condition at a time
• Sequentially in order of baseline stability
• No withdrawal or reversal
• Avoids any associated problems (ethics, changed
  baseline)
• Must have overlap in terms of setting,
  researchers used, participants used, etc. to
  detect potential extraneous variables otherwise
  it is simply a series of stacked AB designs

31
Multiple baseline design
Bob
Homeroom
Minutes on Task
Carol
Homeroom
Critical area
Joe
Homeroom
Sessions
32
Multiple baseline design

• 3 types
• Across participants
• Same settings and behavior, but different
  participants
• Across behaviors
• Same participant and setting, but different
  behaviors
• Across settings
• Same participants and behavior, but different
  settings

33
Across participants
Bob
Homeroom
Minutes on Task
Carol
Homeroom
Joe
Homeroom
Sessions
34
Across behaviors
Min. on Task
Homeroom
Bob
Min. In Seat
Homeroom
Min. w/ No Talkouts
Homeroom
Sessions
35
Across Settings
Homeroom
Bob
Minutes on Task
Math
Bob
Reading
Bob
Sessions
36
1-13 Know the critical features of the changing
criterion design

• .
• Critical features
• Uses the same, behavior, participant, and setting
• Baseline phase
• Shows experimental control by demonstrating that
  changes in the DV occur only when the criterion
  changes

37
Changing criterion design
38
Changing criterion design

• Considerations
• Length of phase
• Considerable variation for control purposes
• Size of incremental change
• Considerable variation for control purposes
• of phases
• More better

39
1-14 Know how single subject designs are most
commonly analyzed.of SSD

• Visual analysis most common
• Assess trends and levels between adjacent phases
• Level value of DV
• Trend direction of the level of DV
• Increasing, decreasing, stable, variable

40
Some useful links.

• http//www.baam.emich.edu/baamessentials/baamsingl
  esubject.htm
• http//www.practicalpress.net/updatenov05/SingleSu
  bject.html