Flashcards in Week 4 Deck (60)
What Type of Research Design is Quasi-Experimental?
Mostly Experimental, some Exploratory
Quasi Experimental is Similar to Experimental Design, except...
it lacks random assignment or comparison group or both
What type of Designs are Quasi Experimental?
One-Group or Multigroup
Another name for one group design is...
Describe one group design
you have a sample study that undergoes a pretest, then an intervention, followed by a post test, and no comparison
Describe a oneway repeated measure design
You have a study sample that undergoes a pretest, then an intervention, followed by either multiple posttests OR a posttest 1, intervention, posttest 2
What are One-Way and One-Group tests vulnerable to?
threats to internal validity
When are One-Group and One-Way Designs defendable?
When there is known natural history (prev research) and the follow up time frame is very short
What is a Time Series Design?
Multiple pretests and posttests act as a pseudocontrol condition
Do One-Group or Time Series Designs have better internal validity?
Describe the Nonequivalent Pretest-Posttest Design
It includes a control group and pre-test, post-test
How is the equivalency of groups measured in nonequivalent pretest-posttest design?
Equivalency is based on prescores for nonequivalent pre-post-test
Describe a Nonequivalent Posttest-only Design
It includes control group, but no pre-test so we have no knowledge of where subjects started
How do you test equivalancy of Nonequivalent posttest-only design?
There is no way to test equivalency of groups based on prescores bc there is no pretest
What type of reseach is Single-Subject Designs?
Mostly Experimental, and somewhat Exploratory.
Another name for Single-Subject Designs
N of 1, N=1
What are N of 1 studies similar to? How?
time series design studies, Rigorous planning (reliable and valid outcome measures) and control phases
Is a Single-Subject Design Prospective or Retrospective?
What is the IV in Single-Subject Design?
What are the levels to the IV in Single-Subject Design?
Phases of Treatment
What is the DV in Single-Subject Design?
Patient Response (aka Target Behavior)
How do Single Subject Design Studies differ from Case Studies/Series?
N of 1 has repeated measures and design phases
In N=1 Studies, What are you evaluating during the baseline phase?
Stability and Trend
How do you observe Stability in N=1?
Observe little variability and if there is a large baseline variability then you may need to obtain more measurements
How do you observe Trend in N=1?
Indicate change is occurring w/o intervention and evaluate whether the rate of change is stable or unstable
What long are phases in N=1?
Can be flexible, but has to be at least until stability is reached, a minimum of 3-4 data points for each phase. Best to have equal phases and interval lengths
What type of outcome measures are used in N=1?
Patient focused and subjective... needs to be quantitative and provide proof of reliability.
Examples of N=1 outcome measures
NPRS, time sitting before pain, gait frequency during 100m walk, Oswestry Disability score, Berg Balance Test
Disadvantage of Single Subject A B Design?
Unsure if B was due to some confounding or maturation effect
Advantages of Single Subject ABA design?
Can determine if A truly caused B to occur since target behavior should return to baseline
Disadvantages of ABA Design?
Behavior must be reversible... ethical issues with withdrawal of treatment
Data Analysis of N=1?
Generally either based on visual trends or variability assessment
How do you know if a data point change occurred by chance or not in N=1?
If it is within 2 SD of the mean then it could have occurred by chance, but if it is outside of 2 SD then it is highly unlikely that the change occurred due to chance.
What Types of Studies are Exploratory?
Cohort Studies and Case-Control Studies, Predictive Reseach
How are data points collected in Observational Studies?
As they naturally exist with no manipulation of variables
What is Exploratory Research?
Systematic investigation of relationships among 2 or more variables (cause and effect)
Longitudinal Research can either be...
Prospective or Retrospective
Prospective or Retrospective, which has more control?
Prospective or Retrospective, which is easier to record reviews?
Correlation and Predictive Studies do not imply...
What are correlation and predictive studies based on?
Covariation in data
What are case-control and cohort studies intended to study?
Risk Factors, association between disease and exposure
What is the difference b/w Case Control and Cohort Studies?
Case Control: Identify subjects based on if they have the disease or not, it is retrospective, works well for rare conditions
Cohort Studies: Identify subjects based on whether or not they were exposed, it is prospective, does not work well for rare conditions
What do Observational studies mostly look at?
Relationships, not cause and effect
What are the 5 criteria to provide evidence for cause and effect in an Observational Study?
1. Time sequence (cause happens before effect)
2. Strength of the Association (need to be strong)
3. Biologic Credibility
4. Consistency (with other studies)
5. Dose-response relationship (the more exposure, the more likely they have the disease)
What is a secondary analysis?
Uses existing database to examine relationship b/w variables.... secondary analysis of project with other main focus, can combine data sets, "big data" projects
What Types of Research are Mostly Descriptive?
Natural History, Normative Research, Qualitative Research, Descriptive Surveys, and Case Studies
Which types of research are sometimes descriptive?
secondary analysis, surveys, and questionnaires
What is the difference b/w random selection and descriptive statistics?
Sometimes w/ descriptive research, you don't have to make inferences back to a population
Examples of Natural History Research?
Adhesive Capsulitis History, Alzheimer's disease...
Examples of Normative Studies?
Normal ROM, Grip Strength, etc... trying to document what is "normal"
What is a case study?
A single case report or case series
How is a case study different from a single subject design?
- A case study is usually retrospective,
- Less standardized and controlled,
- Does not meet IRB definition of research,
- Less internal validity,
- Less external validity
How is a case study less standardized/controlled?
no repeat baseline measurements, assessment time frames are naturally occurring
What is the IRB definition of research?
a systematic investigation designed to develop or contribute to generalizable knowledge
Why are Case Studies a valuable source of information for EBP?
- It shows details vs just averages
- Shows treatment and response in each visit
What are the Purposes of Case Studies?
1. understanding unusual patient conditions
2. examples of innovative or creative therapies
3. generating and testing theory
4. providing future research directives