Week 4 Flashcards
External validity
Whether research results can be generalised to the context we are trying to make inferences about
Prospective vs retrospective
prospective: baseline research and then follow up (predictive qualities)
Retrospective: One measurement
Cross sectional vs Longitudinal
cross-sectional: measure one cohort at one time
longitudinal: measure a cohort at different intervals
Self-report vs Observed
self report: asking participants to report on measure i.e. food diary or memory
observed: researcher/observer interactions or review credit card data to measure
Replication Crisis
Where published experiments may not replicate. Of 100 studies reviewed, only 30% found significant effects and the effects were half the size initially reported.
Multiple reasons behind this i.e. questionable research practices and publication bias are two.
How to fix the replication crisis?
Pre-register study designs
Open data sharing
Change research incentives
Randomised Control Trials (RCTs)
Involves randomising participants either individually or by group
Is the gold standard for demonstrating causality and measuring the effect of the intervention
Systematic and meta analytical reviews
Involves systematically finding all other studies and considering them together
Present a synthesis of information
Considered top of the evidence hierarchy
Example: can computer delivered interventions reduce alcohol intervention
Correlational research
Involves measuring two variables to see if they are related
Correlations can be + and -
Is the +/- change in one variable associated with a change in the other
Qualitative research
Analysis of participant reviews to provide an understanding of a certain condition/disease
Data science
produces insights from data using scientific analytical methods
Health economics
branch of economics designed to help decision making in healthcare
