Related Experiment Videos
Predicting smoking behaviour in census output areas across Scotland
Jamie Pearce1, Paul Boyle, Robin Flowerdew
1School of Geography and Geosciences, University of St Andrews, St Andrews, Fife KY16 9AL, Scotland, UK. jrp6@st-andrews.ac.uk
Health & Place
|May 20, 2003
Summary
This study introduces a new method to estimate smoking probability across Scotland, providing detailed geographical insights into smoking behaviour for various age and sex groups. These findings aid in understanding disease variations linked to smoking patterns.
Area of Science:
- Public Health
- Geographic Epidemiology
- Statistical Modeling
Background:
- Spatially disaggregated data on smoking behavior is scarce, hindering accurate geographical estimations of smoking incidence.
- Understanding the geographical distribution of smoking is crucial for public health research, particularly for diseases like lung cancer.
Purpose of the Study:
- To develop a novel technique for estimating smoking probability across small areas in Scotland for different age and sex demographics.
- To provide geographically detailed smoking estimates that can inform public health strategies and epidemiological studies.
Main Methods:
- Utilized individual-level data from the Scottish Household Survey on smoking behavior.
- Combined survey data with 1991 census measures (output area and pseudo-postcode sector) to model smoking probability.
- Employed statistical modeling to predict smoking prevalence by age and sex for output areas throughout Scotland.
Main Results:
- Successfully developed a method to estimate smoking probability at a geographically granular level across Scotland.
- Generated detailed smoking predictions by age and sex for small output areas.
- This represents the first instance of such fine-grained geographical smoking estimates for Scotland.
Conclusions:
- The developed technique provides unprecedented geographically detailed estimates of smoking probability in Scotland.
- These estimates are valuable for public health planning and for controlling confounding factors in epidemiological studies of smoking-related diseases.