(go to Outline)
Sampling bias results from not selecting a truly random sample which is representative of the larger population. For example:
- The chiefs in some villages insisted that the survey team weight and measure their many children. The survey teams did this and included these data in the survey.
- This would probably lead to an underestimate of the prevalence of malnutrition because the chief's children are probably better off than children in general.
- When making lists of households in several villages, the survey team forgot to include households where an ethnic minority lived. People in this minority are routinely discriminated against and cannot, in general, use the villages borehole, but must drink from nearby ponds or streams.
- This would lead to an underestimate of the prevalence of recent diarrhoea as people, especially children, in the ethnic minority are probably more exposed to diarrhoea-causing organisms.