PLEASE NOTE:
We are currently in the process of updating this chapter and we appreciate your patience whilst this is being completed.
This set of notes has already looked at the use of statistical methods in epidemiology, for example, calculation of:
- Prevalence and incidence
- Measures of effect and population impact
- Outcome measures for specific trial designs
Additionally, see 1b Statistical Methods module of the website which includes detailed descriptions of the commonly used statistical techniques that feature on the DFPH syllabus. Epidemiology and the role of chance1 It is important to remember that epidemiological studies are limited to a sample of individuals from a population, because it is usually impractical to include everyone. Estimates from the sample group are used to make inferences about the wider population. However, observed estimates from sub-samples of a population may differ from the estimate gained if another sub-sample of the same population had been used. Statistical methods can be used to assess the probability of obtaining an observed estimate from a sample by chance alone, and to assess the range of values within which the actual population estimate is likely to lie. Key techniques used to do this are:
- p-values are obtained from statistical significance tests. They give the probability of obtaining the observed estimate by chance alone. Conventionally, if the p-value is <0.05 (i.e. the probability of obtaining the observed estimate by chance alone is 0.05, or 1/20) the role of chance is rejected and the observed estimate is taken to be a significant estimate for the association under investigation.
- Confidence intervals tell us the range in which the true population estimate is likely to fall.
These techniques are also covered in more detail in the statistics section of this website. References 1. Bailey L, Vardulaki K, Langham J, Chandramohan D. Introduction to Epidemiology. Open University Press, 2005.
© Helen Barratt 2009