Year 1 Flashcards
Draw the picture
Case report case series
Quantitative
Observational
Descriptive
Detailed description to generate hypothesis
+ first clues of new diseases
- no proof no comparison
Cross sectional prevalence
Quantitative
Observational
Descriptive
Snapshot in time
+ quick and cheap
- cant confirm causation
Ecological
Compares exposure with another outcome
+ low cost
- ecological fallacy
Case control
Quantitative
Observational
Analytic
Compares cases with controls with history of exposure
+ investigates multiple causal exposures
- susceptible to bias
Cohort
Quantitative
Observational
Analytic
focuses on exposure then outcome
+ measures incidence
- expensive and time consuming
Cross sectional analytic
Quantitative
Observational
Analytic
Uses odds ratio, snapshot
+ quick and cheap
- single time point
Explanatory RCT
Quantitative
Experimental
randomised control trial
Highly controlled conditions, select participants, perfect
+ removes bias
- expensive and time consuming
Pragmatic RCT
Quantitative
Experimental
Randomised control trial
More flexible setting, real world
+ removes bias
- expensive and time consuming
Qualitative
Non numerical data
Splits into in depth interviews, focus groups and ethnographic studies
Descriptive
Case report
Case series
Cross sectional prevalence
Longitudinal
Ecological
Analytic
Case control
Cohort
Cross sectional analytic
Bradford hill criteria
Analogy
Coherent
Consistency
Experimental evidence
Plausibility
Temporal relationship
Susceptibility
Strength of association
Spurious
Variables correlated without relation
3rd variable problems
3rd problem linking other 2 variables
Anscombes quartet
Statistical relation doesn’t equal distribution
Colliders
Exposure and outcome independently influence 3RD variable
Can obscure real/reveal false associations
Selection bias
Over or under representation of groups of participants
Observation bias
Participants modify behaviour when they’re being watched
Confirmation bias
Look for specific outcome
Publication bias
P hacking
Number manipulation
Attrition
Unequal loss of participants