Chapter 9: Quantitative Research Basics Flashcards
_____ is a systematic and principled way of obtaining information; a structured process of investigating facts and theories.
Research
Why is it important to do research in the health professions? There are 4 reasons.
- It is unethical not to 2. It contributes to preserving the worth, dignity, and respect of clients 3. Natural curiosity and quest for excellence drive research 4. Some people find research exhilarating
Why is research important for OT? 5 reasons
- Professions research obligation 2. Necessity of research for professional recognition and support 3. Generating foundational knowledge 4. Providing evidence of the need for OT services 5. Testing assessments used in therapy
_____ research places an emphasis on objectivity; measurable; deductive; pre-determined structure
Quantitative
______ research is subjective; perceived/perceptions/feelings; inductive; process-oriented
Qualitative
What characteristic of research is described by how procedural/protocol driven the research is?
Rigor
What characteristic of research is described when you ask if its really based on good data?
Empiricism
What characteristic of research is described by asking if the research makes sense?
Logic
What characteristic of research is described by asking if the research can be applied to many populations?
Communality
This research term used to describe when you “do something with the variables”
Manipulation
This research term is used for when a group does not get the same manipulation as the other group
Control
This part of research helps to control bias. There can be ____ sampling and _____ assignment
Randomization
Types of control: This type of control is used when there is an A group or a B group, etc.
Alternative Intervention
Types of control: When one group is given a “fake” intervention
Placebo
Types of control: “When is the effect of a specialized program in OT?”
Standard methods of care
Types of control: give one group treatment later on; use them as a control group in the beginning
Wait-list
What phase of Full Clinical Trials does this describe? Initial development of drug or therapy; designed for safety and tolerance/ determining optimal dose; small scale studies using simple designs; focus on developing the best possible and safest treatment
Phase I
What phase of Full Clinical Trials does this describe? Preliminary evidence of treatment effectiveness; assess feasibility of launching rigorous test; pilot test of treatment; look for signs of side effects; identify refinements to improve the intervention; may start publishing; usually missing randomization (essentially a pilot test of the treatment)
Phase II
What phase of Full Clinical Trials does this describe? Full test of treatment; RCT standard design used at this phase; develop evidence about treatment efficacy; has to be randomized; this is what journals want to see; often involve large and heterogeneous sample of participants
Phase III
What phase of Full Clinical Trials does this describe? effectiveness of intervention in general population; focus on post approval safety surveillance; focus on long-term consequences over larger population; intervention given to everyone in general population
Phase IV
What are some strengths of Experimental design? (3)
- Strongest evidence for intervention effects 2. Least bias 3. Causal Relationships
What are some weaknesses of experimental design? (4)
- impractical 2. limited control in clinical settings 3. hawthorne effect (people change behavior because they know they are in a study) 4. hard to control all the variables in a clinic
This type of quantitative research is not randomized
Quasi-experimental research design
Strengths of quasi-experimental research (2)
- more practical for clinical settings 2. more acceptable to participants
Weaknesses of quasi-experimental research (3)
- cause and effect inferences are less compelling 2. considered weaker designs 3. less publishable
This type of quantitative research has no intervention
Non-experimental research design
What is this non-experimental research design? Relationship; association; researcher has no control over variables; different designs have different degrees of evidence; no independent/dependent variable
Correlational
What is this non-experimental research design? Phenomenon existed in present is linked to the past; begins with dependent variable and examines relationship with independent variable
Retrospective Ex. looking back in medical records
What is this non-experimental research tool? data on both variables collected at one point in time; not considered a design but a method of data collection
Cross-sectional
What is this non-experimental research design? Start with presumed cause and then go forward to presumed effect
Prospective
What is this non-experimental research design? observe, describe, and document aspects of natural occurring experience; starting point for hypothesis generation; two types (univariate, correlational)
Descriptive non-experimental research
This type of descriptive study uses incidence and prevalence studies
Univariate
Proportion of total population who have a particular health related condition
Prevalence
How many people have onset of condition during a given time span
Incidence
This type of descriptive study observes, describes, and documents aspects of natural occurring experience; describe relationships among variables; does not support inferences of causality (can’t say that x causes z)
Correlational
Strengths of correlational research (3)
- Efficient at collecting large amount of data 2. Plays a role in developing evidence base for causal inference later 3. Strong in realism
Weaknesses of correlational research (3)
- Considered weakest research designs 2. Difficulty supporting causal inferences 3. Selection bias
Basic posttest-only design
Basic pretest-posttest design with optional repeated follow-ups
(could have an O4 as well)
Multiple Intervention Design
Wait-list
(delay of treatment)
Crossover design (participants serve as their own controls)
Nonequivalent control group, posttest only design
(Kelli called it quasi-experimental post-test so not 100% about this one)
One group pretest-posttest design
Time series design
Time series nonequivalent control group design
Non-equivalent control group pretest-posttest design
A true randomized control trial has what 3 characteristics?
Manipulation, Control, and Randomization
Occurs when researchers take steps to ensure subgroups of participants are allocated equally to conditions (ex. men and women randomly assigned to conditions separately, thus ensuring that both men and women equally received the interventions)
Stratification
Bias that results from participants expectations of efficacy of treatment
Could be either placebo OR particpation expectation bias
What are some examples of survey research?
census is a survey that covers entire population; many surveys just use a sample (that’s a sample survey)
most respected method: personal interview
other examples: telephone interviews, questionnaires
direct attempt to see if findings obtained in a study can be duplicated in another study
replication studies