Question

Salmonella in the Caribbean. A case study

Any other factors such as power or attrition that may affect the

            1/ validity

            2/ reliability

            3/ generalizability of the population surveilled?

Answer 1

When data for a research is collected for a prolonged period of time, some of the study subjects may drop out and the difference in the size of population through out study may affect the validity, reliability and generalizability of outcomes. Attrition mainly occurs in longitudinal research.Difference in characteristics of experimental and control group as well as in the follow up data affects the research validity. Attrition bias leads to incomplete or missing data that may affect the internal and external validity It may cause incorrect correlation and compromise the reliability of outcome. The results cannot be generalized when the external validity is weak. Though there are no steadfast rules regarding susceptibility of a study to bias in relation to attrition,experts suggests that an attrition rate of less than 5% is of less concern and that of more than 20% create major concern.