Research Questions & Design Flashcards
What is the 8 criteria for quantitative research?
- Define the problem 2. Develop the research question a) Developing specific research objectives (aims) b) Developing hypotheses Create a research design 4. Choose methods a) Sample b) Measures 5. Selecting analytical methods 6. Implementation 7. Analysis & Interpretation 8. Reporting
What are the two types of reasoning when defining the problem?
Deductive Reasoning = From generic to specific eg. All graduate students hate Statistics. You all are Kinesiology graduate students. Therefore, you all hate Statistics Inductive Reasoning =From specific to generic ou all hate Statistics. eg.Since you are all graduate students, then all graduate students must hate Statistics.
What is the criteria when defining a research question? ( (step 2, acronym)
F.I.N.E.R Criteria Feasible - time, access to sample, expertise, $$$ Interesting - to discipline, field, society Novel - confirms theory/model, refutes theory/model, adds to existing knowledge Ethical - Risk vs. Reward Relevant -scientific knowledge, academia, clinical translation
WHat is the acronym used to frame your research question?
PICOT Population (sample, accessible, target) Intervention/Exposure Comparison Outcome (dependant variable) (Timing)
A _____ or _____does not have to involve an active manipulation but a _____ does
intervention (or exposure), control Control: Cause vs effect,.(control for (all) extraneous variables and systematically manipulate the independent variable) vs. Casual relationship
What are the 3 types of research questions and give examples
Difference Research Question - Is the outcome of interest different depending on the intervention (exposure)/comparator of interest within the population of interest? (Exposure/intervention &control) Association Research Question - Is there an association between the exposure of interest and the outcome within the population of interest? (exposure only) Descriptive Research Question - How frequent (or severe) is the outcome of interest in the population of interest? (No exposure/intervention)
What are the two types of hypotheses?
Directional research hypothesis
Predicts positive or negative relationship,
Non-directional research hypothesis
Predicts relationship, change, or difference but not specify direction
SLIDE 33, matching the appropriate research design based on the question
SLIDE 33, matching the appropriate research design based on the question
What is the acronym for when choosing research design?
PIRCP (People in Russia Cant Poop)
Prevalence = Cross-sectional
Incidence = Cohort
Risk factor = Cohort
Causation = RCT/Experimental
Prognosis = Cohort
Explain the differences between qualitative and quantitative research design
What is validity and what are the two types?
Do your results truly represent the phenomenon your claim to measure?
Internal Validity
- Confidence that outcome observed in dependent variable was due to independent variable
External Validity
- Extent to which your findings generalize to your population of interest (other groups or contexts)
What are the 3 types of Observational Designs?
Cross-sectional, cohort, case-control
What is cross-sectional, what is its uses, advantages, and disadvantages
- Recruitment is based upon population
- Each individual represents a snapshot in time
Uses: descriptive, prevalence, association, difference research questions
Advantages: Relatively quick & inexpensive
Disadvantages: Weak evidence for causation as time-dependency is unknown (don’t know which came first), confounding
What is a cohort study, what is its uses, advantages, and disadvantages
- Recruited is based on population
- Measured longitudinally (over time)
Uses: Incidence, prognosis, identification of risk factors
Advantage: Demonstrates time- dependency, can infer population relative risk, understand change over time
Disadvantages: Poor detection of rare diseases, confounding
What is a case-control study, what is its uses, advantages, and disadvantages
- Recruited by the outcome (case/not case)
- Measured cross-sectionally or longitudinally
Uses: association, differences research questions (risk factors, outcomes)
Advantage: High concentration of cases (good for rare diseases)
Disadvantage:
Sample (selection) bias Recall bias (cross-sectional) Confounding
What is a confounding variable?
A variable that influences both the dependent variable and independent variable, causing a spurious association
What are the 3 types of experimentl designs and exampls of each?
True Experimental designs (2)
- post-test only
- pre-post test w/ randomization
Quasi Experimental Designs (3)
- pre-post test – non-random assignment
- repeated measures design (2 types)
Pre Experimental Designs (3)
- one group pre and post test
- one-shot design
- static group comparison design
Characteristics of the true experimental design (post-test only, pre-post test w/ randomization)
True experiment: Have both a Control group & Random Assignment to groups
Characteristics of the quasi-experimental designs (pre-post test - non-random design, repeated measures [individual serves as a control], repeated measures [Multiple pre-post measures] )
They all have either a Control group OR Random Assignment to groups (not both)
Characteristics of the pre-experimental designs (one group pre and post-test, one-shot design, static group comparison)
Characterized by absence of a true Control group & Random Assignment to groups
For static, since the control group is random:
- Control group is not actually controlling, waste of resources
