cross sectional and ecological studies Flashcards
2 types of study
descriptive- no hypothesis (eg proportion of obese people in pop.)
analytic- hypothesis driven, analyse associations (eg is proximity to fast food joints associated with obesity)
CS
how is information collected
observational study, gather information once or at one time period (repeat studies may become longitudinal)
exposure and outcome measured simultaneously
CS
what part of population is used
whole pop. or a representative sample with no follow ups
CS
descriptive vs analytic
D- freq. or distribution of exposure at one point (generates hypotheses)
A- compare exposed and non exposed to examine association involves hypotheses testing to see if its significant
CS
where is it used
health service planning resource allocation and monitoring population health
CS disadvantages
can’t prove causation/correlation, not suitable for short, rare, highly fatal diseases, also can be subject to bias
CS advantages
quick cheap easy, large pop., can study multiple exposures/outcomes simultaneously, good to measure prevalence
CS
repeat studies
eg health surveys, can make time series
E
observational
based on groups eg country regions or school
E
what’s it used for
study occurrence in groups or associations of groups (eg GDP and diabetes)
E
3 variables used to capture characteristics
aggregated - summarised individual data
environmental- physical environment eg temp./pollu.
global- everyone in group has same level of exposure eg presence of cycle paths
E
descriptive vs analytic
D- hypothesis generating describes distribution and patterns of outcome
A- hypothesis testing examines relationship in different groups or population over time
E
studies can be experimental but also susceptible to
ecological fallacy where incorrect inferences are made about individuals based on populations data
E advantages
easy to use with data and quick and cheap, ideal for population metrics of health eg life expectancy
E disadvantages
ecological fallacy and other bias, sometimes to standardisation to ensure accurate comparisons