Question 5: Why does SPSS put the words 'For cohort' in the Risk Estimate table?
To calculate the relative risk of cough given a history of bronchitis, we need to calculate the risk of cough for children with a history of bronchitis and the risk of cough for children without a history of bronchitis. We then divide one by the other. We can only calculate these risks if we start with a group of people with a history and another group without a history, i.e. a cohort of people followed over time or a cross-sectional study where we obtain both current symptoms and history from the same sample. We cannot do it if we start with a group of children with cough and a second group of children who do not have cough, then find out whether they had a history of bronchitis. This would be a case-control study.
The relative risk is only interpretable if we have a cohort study, cross-sectional study, or clinical trial, not if we have a case-control study. The authors of SPSS are reminding us of this.
Having said that, the output is pretty confusing.
In some circumstances, we can estimate the relative risk in a case-control study using the odds ratio, but that is beyond the scope of this module.
Back to Exercise: Cough during the day or at night and history of bronchitis.
To Applied Biostatistics index.
To Martin Bland's M.Sc. index.
To Martin Bland's home page.
This page maintained by Martin Bland.
Last updated: 21 November, 2006.
Back to top.