Evaluating Research Flashcards
(32 cards)
What does PICOT stand for?
Population, Intervention, Control, Outcome, Time.
Population
Who you’re studying and why.
Population subsets.
Target population: General or specifc
Accessible population: People available to study
Sample: The people you do study
How can you characterize populations?
Demographics: Age, sex, identity, nationality, ethnicity, etc.
History: Diagnosis, condition, lifestyle, training, risk factors etc.
What is intervention
What you’re doing to change them (participants, or what the study is trying to test.
What is manipulated interventions
The researchers artificially produce conditions, can be done by pharmaceuticals, training, posture, gait.
What is non-manipulated interventions
Things researchers cannot control, conditions or groups naturally exist. Examples, clinical diagnosis, age, existing lifestyle, gender, occupation.
What is Control
Want to see how the intervention affected person/people. Done by comparing the intervention to: no intervention, placebo intervention, sham intervention, standard intervention, before or after the intervention.
What does the combination of intervention(s) and control(s) describe?
Independent variables
What is Outcome?
What you’re measuring (discrete/categorical, continuous), aka dependent variables.
What is Time in the context of research?
How long you are tracking the outcome.
What is a cross-sectional study?
Assess at a “single” time point. Assumes no development between intervention/control. Usually over a short amount of time.
What is a longitudinal study?
Track over time to allow progression. Explicit development. Over longer periods of time like an 8 week exercise intervention.
3 things to consider when writing a hypothesis:
- Refer to a specific set of intervention/control and outcomes.
- Indicate direction of changes whenever possible.
- Explicitly tie rationale to theory/evidence.
What are control variables?
Measured things that keep consistent as independent variable(s) change. Done to ensure only the independent variable affects the dependent variable.
What are confounding variables?
Unmeasured things that could also affect dependent variable(s). Sources of error (control variable you missed).
What is reliability?
How consistent are outcomes/measures for identical inputs/attempts? Can be expressed as an error term (units %).
What is validity?
How well does an instrument measure what it’s intended to measure? Internal, external, and others.
Can something be reliable but not accurate?
Yes.
What is internal validity?
The degree to which the independent variable causes changes in the dependent variable.
How does Selection threaten internal validity?
The groups might not be similar enough. Solution: randomize.
How does History threaten internal validity?
An outside event altered dependent variables. Solution: Randomize.
How does Maturation threaten internal validity?
Participants naturally change during experiment. Solution: Randomize, controls.
How does testing threaten internal validity?
Learning tests/tasks. Solution: randomize, controls.
