The Significance of Selection in Lung Cancer Studies
Author Information
Author(s): R. Korteweg
Primary Institution: null
Hypothesis
Can the selection bias in prospective investigations explain the association between smoking and lung cancer?
Conclusion
The study suggests that while selection bias affects death rates for smokers, a significant portion of the excess death rates from lung cancer and coronary disease is likely due to other factors.
Supporting Evidence
- The study indicates that the death rates found by Hammond and Horn are lower than official statistics, suggesting selection bias.
- Berkson's model shows that the reluctance of smokers to participate can distort death rate findings.
- The research concludes that a significant part of the excess death rates for smokers is not due to selection bias but to other factors.
Takeaway
This study looks at how the way we choose people for research can change the results about smoking and lung cancer. It finds that some of the extra deaths from lung cancer in smokers might not just be because they smoke.
Methodology
The study uses a model to analyze the effects of selection bias on death rates in smokers compared to non-smokers.
Potential Biases
There is a risk of selection bias due to the psychological reluctance of smokers to participate in studies that may highlight the dangers of smoking.
Limitations
The reference population may not be representative of the entire American population, and the assumptions made about cooperation rates among smokers may not hold true.
Participant Demographics
The study focuses on 187,766 American white males aged 50 to 69.
Want to read the original?
Access the complete publication on the publisher's website