Singel Subject Design Flashcards
Hierarchy of Scientific Evidence:
1 (toppen av pyramiden) - Meta-Analysis and Systematic Review
2 - Randomized controlled trials.
3 - Cohort Studies.
4 - Case-Control studies.
5 - Cross Sectional studies.
6 - Animal trials and in vitro studies.
7 - Case report, opinion papers, and letters.
Single-case research design:
The phenomena of interest are studied using a single subject or a small group of research subjects (also referred to as N-of-1 designs or single-subject research).
The 3 Major types of single-case research:
1) Case studies
2) Quasi experimental designs
3) Experimental designs
1) Case studies:
- Refers to a broad array of approaches, most often used for observing an individual and reporting on their interactions with variables of interest.
- Unit case studies : Individual, Family, Class
Advantages of case studies:
- do not require control conditions or comparison group.
- can be easily incorporated into routine clinical work.
- offer clinicians both a useful research tool as well as a source of data directly relevant to clinical practice.
- valuable in the study of rare phenomena.
- provide opportunity to observe the behavior of interest “as it naturally occurs”, which may be useful for generating a hypothesis.
- provide an incredibly efficient and flexible tool in the assessment and treatment development and evaluation process.
Disadvantages of case studies:
- Lack the methodological requirements to draw valid inferences about relations among variables.
- Typically do not include design features such as objective assessment, systematic data collection or analysis, specified manipulation of the independent variable, or replication of treatment effects.
Clinical Research:
How do I help the most people with these kinds of problem?
Clinical Practice:
How do I help this person sitting in front of me right now?
Single case studies:
- Avoids small, unimportant effects
- Links science to practice, and practice to science.
- Facilitates innovation.
- Fits into clinical settings.
- Creative and flexible.
Single case design essentials:
- Internal validity - effect because or intervention? -> adequate comparison conditions.
- External validity - Generalize? -> replicate, replicate, replicate.
This requires - repeated continues measurements and systematic manipulation of intervention.
2) Single-case Quasi-experimental design:
- Repeated assessment of the dependent variable -> reliable and valid measures, should occur early, frequently, and consistently , should be feasible.
- Intervention -> independent variable, can be behavioral, cognitive, psychodynamic, or other methods, intervention should be well specified.
- Choose intervention targets that are: Frequent, concrete and quantifiable, stable without treatment.
3) Singe-case Experimental design:
- Includes a controlled variation of the independent variable and measurement of the effects of this variation allowing the experimenter to examine causal relations between intervention and outcome.
- Design essentials -> Multiple phases and conditions, Replication of experimental effects.
Multiple phases and conditions:
- Each individual participates in both treatment and control conditions.
- The vital element is the inclusion of multiple conditions (that it, >1) for each subject.
- A=non-treatment phase
- B=treatment phase
- Examples: ABA or ABAB.
Replication of experimental effects:
- Replication of experimental effects should occur both within and between studies.
- This is one of the greatest strengths of such approaches.
Primary Research Design:
- Reversal design
- Multiple-Baseline Designs
- Changing Criterion Designs
Reversal Design:
- ABAB -> behavior first assessed during a baseline(A1) then during experimental phase, in which intervention/manipulation is administered (B1), followed by a return to the baseline phase (A2) during which the behavior is expected to reverse to its baseline level, and so on.
- ABAC -> In cases in which the specified intervention was not effective ( no behavior change), researcher may wish to test the effect of adding an additional treatment component or a modified version of the intervention.
Multiple-Baseline Design:
- Experimental condition is introduced in temporal sequence to different behaviors.
- Typically only one baseline (no reversal) and one intervention condition.
- power of such designs comes from demonstrating that change occurs when, and only when, the intervention is directed at the behavior in question.
Mutiple-Baseline Design:
- Advantages -> Can demonstrate causality by replicating the effects across different behaviors. It does not require the withdrawal of effective intervention.
- Limitations -> Convincing demonstration of causality requires that change does not occur across behaviors until the intervention is applied. Therefore, interventions that lead to change that generalize across all areas will limit the strength of the inferences that can be drawn.
Data analysis:
- Statistical analysis
- Visual inspection
Single-Subject Research:
- Internal validity -> Always a concern in single subject designs.
- Two Major threats -> Instrumentation and Specificity of variables
- External validity -> Lack of this is a major concern with single-subject designs
- Generalizability -> addressed throughout multiple replications of the same treatment and design that produce similar results for a number of different participants.
