The use of epidemiological tools in conflict-affected populations: open-access educational resources for policy-makers

Table of Contents

1. Welcome
2. Introduction: Epidemiology in crises
3. Ethical issues in data collection
4. Need for epidemiologic competence
5. Surveys - Introduction
6. Surveys - Description of sampling methods
7. Surveys - Sampling error, bias, accuracy, precision, & sample size
   1. Bias and sampling error
   2. Bias and sampling error 2
   3. Bias
   4. Measurement bias
   5. Sampling bias
   6. Sampling error
   7. Bias and sampling error - Quiz
   8. Confidence intervals
   9. Confidence intervals - Quiz
   10. Accuracy and precision - theory
   11. Sample size
8. Surveys - Resources required for surveys
9. Surveys - Critiquing survey reports
10. Surveillance - When to do surveillance
11. Surveillance - Methods
12. Surveillance - Common problems
13. Programme data
14. Rapid assessment
15. Mortality - Indicators and their measurement
16. Mortality - Data sources
17. Mortality - Interpretation and action
18. Nutrition - Introduction and background
19. Nutrition - Indicators and their measurement
20. Nutrition - Data sources
21. Nutrition - Interpretation and action
22. Health services
23. Vaccination programmes
24. Water supply, sanitation, and shelter
25. Violence
26. Presentation of results
27. Formulating conclusions and recommendations
28. Dissemination and action

Bias and sampling error

(go to Outline)

The concepts of bias and sampling error are often confused with each other and poorly understood, even among those who frequently do surveys. This is a shame because it results in much wasted effort and invalid conclusions. The following explanation attempts to clarify the situation.

To being with, the estimate from a survey is never exactly identical to the actual value in the population, even if all the procedures are done correctly. For example, in a hypothetical population in which precisely 50.0% of non-pregnant women of child-bearing age have anaemia, a very well-done survey of 500 women shows that 225 (45.0%) have anaemia.

Why might this be true? (Click here for answer.)