ecological studies Flashcards
unit of analysis
ecological studies
groups
EX: countries, states, regions, neighborhoods, hospitals
we measure
ecological studies
exposure and outcome for each group
relationship is defined by
ecological studies
correlation
correlation
ecological studies
defines relationship btwn 2 variables
-1 to 1
correlation coefficient
quanitfies magnitude of association btwn possible risk factor and outcome
r^2
proportion of variation in outcome explained by variable
r of 0.40 means
r^2 = .16 = 16%
correlation btwn X and Y was 0.40
amount of variation is 16%
limitations
ecological studies
- no individual data
- no info on confounders
- no temporality
- ecologic fallacy
- group data may not measure what you want to study (ex: individual flu exposure)
ecologic fallacy
ecological studies
observations at group level may not represent exposure-disease relationship on individual level
occurs when incorrect inferences about individuals are made from group data
results from ecologic study may be reverse of an individual study
strengths
ecological studies
some exposures vary little within population
we can study a greater range of exposures b/c we compare across populations (ex: dietary fat intake on CHD: study diff countries with diff diets)
unique uses
ecological studies
variables that can only be measured for groups
water fluoridation
health care policies by countries
pros
ecological studies
cheap, convenient
can use previously collected data
individual data may not even exist
can generate hypotheses for other epi studies (cohort, case control)
cons
ecological studies
can’t draw inferences about individual data association
difficult to control for confounders
data previously collected may vary in quality
