Post on 03-Jan-2016
transcript
Educational Research
Chapter 10Single-Subject Experimental
Research
Gay, Mills, and Airasian
Single-Subject Research Topics discussed in this chapter
The unique characteristics of single-subject designs
External and internal validity Designing single-subject research Three types of single subject designs Analyzing data Replication
Single-Subject Research Two unique characteristics
The sample size is one subject Each subject serves as his or her own control
Notation ‘A’ indicates a non-treatment phase usually referred
to as a “baseline” ‘B’ indicates a treatment phase Examples
A-B-A indicates a non-treatment baseline phase followed by a treatment phase which is followed by a return to a non-treatment baseline phase
A-B-A-B indicates four phases of treatment (baseline, treatment, baseline, and treatment)
Objectives 1.1, 1.2, 1.5
Single-Subject Research Why use a single-subject design?
In some situations it is unethical to deny a control group treatment, so a single-subject design is the alternative
Denying Title I resources to children who qualify for them
In other situations there are low incidences of subjects available to participate
Children with specific special needsObjective 1.3
Single-Subject Research External validity
The lack of external validity is the major concern with single-subject designs
Generalizability is addressed through multiple replications of the same treatment and design that produce similar results for a number of different participants
Objective 1.4
Single-Subject Research Internal validity
Internal validity is always a concern with single-subject research
Two major threats Instrumentation Specificity of variables
Controlling threats Baselines are multiple measures of pretest
performance By repeating baseline measures over a period of time
threats to internal validity can be controlled History Maturation
Objectives 1.6 & 1.7
Single-Subject Research
Number of manipulated variables Only a single variable should be
manipulated in single-subject designs Adding and withdrawing more than
one variable becomes problematic in terms of analyzing the individual effect of either variable
Objective 1.8
Single-Subject Designs Three major categories
A-B-A withdraw Alternating phases of baseline (A) and
treatment (B) Alternatives include the A-B-A-B design
Multiple baselines The systematic addition of behaviors,
subjects, or settings for intervention Used when baselines cannot be recovered
after treatment has been receivedObjective 1.9
Single-Subject Designs Three major categories (cont.)
Alternating treatments Rapid alternation of treatments to a
single subject to assess the effectiveness of two or more treatments
Objective 1.9
A-B Withdraw Design The A-B design
O O O O O O O X O X O X O X OBaseline Treatment
Internal validity threats are of concern Use of designs with additional baseline
and/or treatment phases helps to control threats to internal validity
Objective 1.10
A-B-A Withdraw Design The A-B-A design
O O O O X O X O X O O O O OBaseline Treatment Baseline
If the outcome is better during treatment than either baseline, the treatment is likely effective
Internal validity threats can be controlled The major concern
The experiment ends with the subject not receiving the treatment
If the treatment has been shown to be effective this is an ethical concern
Objective 1.10
A-B-A-B Withdraw Design The A-B-A-B design
O O O O X O X O X O O O O O X O X O X OBaseline Treatment Baseline Treatment
Internal validity threats can be controlled The effects of the treatment can be demonstrated twice If the results are the same, it is likely the influence of
extraneous variables has been controlled Ethical concerns related to the A-B-A design are
eliminated The right pattern of results provides convincing
evidence of the effectiveness of the treatment
Objective 1.10
Multiple Baseline Designs Three basic multiple baseline designs
Across behaviors Data are collected on several behaviors for a
single subject Treatment is applied to each behavior one at a
time until all behaviors have been treated Across subjects
Data are collected on several subjects for one behavior
Treatment is applied to each subject one at a time until all subjects have been treated
Objective 2.1
Multiple Baseline Designs Three basic multiple baseline
designs (cont.) Across settings
Data are collected on one behavior for one subject across several settings
Treatment is applied to each behavior one at a time in each setting until all settings have been treated
Objective 2.1
Multiple Baseline Designs Three basic multiple baseline designs (cont.)
An example across behaviors Behavior 1 O O X O X O X O X O X O
Behavior 2 O O O O X O X O X O X OBehavior 3 O O O O O O X O X O X O
Design concerns If behaviors are treated the behaviors must be
independent of one another If subjects are treated the subjects must be similar If settings are treated the settings must be as natural
as possible
Objective 2.1
Multiple Baseline Designs Advantages
Can be used when baseline data are not recoverable after treatment
The effects of reinforcement are designed to be maintained after the reinforcement is removed
Can be combined with A-B-A designs to establish a very convincing case for cause and effect
An A-B-A design is applied across three behaviors
Objective 2.2
Alternating Treatments Design Involves the relatively rapid alternation of
treatments for a single subject Treatment does not occur at fixed time periods Treatments are changed sporadically
Advantages Useful in assessing the relative effectiveness of two
or more treatments No withdraw of treatment is necessary No baseline is needed The effects of treatment can be studied quickly and
efficientlyObjectives 3.1 & 3.2
Data Analysis and Interpretation
Two phases An evaluation of the adequacy of the
design An assessment of the treatment
effectiveness Two techniques
Visual inspection of the data Graphical presentation of the results
Objective 4.1
Data Analysis and Interpretation Significance
Clinical significance Effects related to the behaviors being treated Meaningful in a “real” sense Parallels the importance of practical significance
Statistical significance Statistical tests are available but often result in
statistical significance that has little if any clinical significance
Use is currently debated in the field
Objective 4.2
Replication Replication is an important aspect of
single-subject research The more one’s results are replicated the
more confidence one has in the procedures that produced the results
Three stages of replication Direct replication
Same researcher, same subjects, specific setting Simultaneous replication refers to the use of
different subjects and increases the generalizability of the study
Objectives 5.1 & 5.2
Replication Three stages of replication (cont.)
Systematic replication Follows direct replication efforts Involves different researchers, behaviors, or
settings Over time techniques are identified that
consistently produce effective results Clinical replication
Follows systematic replication Involves the development of treatment packages
composed of two or more effective treatmentsObjective 5.2