attrition

this is a closely related concept to drop-out rate, the process by which some participants or cases in research are lost over the duration of the study. For example, in a follow-up study not all participants in the earlier stages can be contacted for a number of reasons – they have changed address, they choose no longer to participate, etc.

The major problem with attrition is when particular kinds of cases or participants leave the study in disproportionate numbers to other types of participants. For example, if a study is based on the list of electors then it is likely that members of transient populations will leave and may not be contactable at their listed address more frequently than members of stable populations. So, for example, as people living in rented accommodation are more likely to move address quickly but, perhaps, have different attitudes and opinions to others, then their greater rate of attrition in follow-up studies will affect the research findings.

Perhaps a more problematic situation is an experiment (e.g. such as a study of the effect of a particular sort of therapy) in which dropout from treatment may be affected by the nature of the treatment so, possibly, many more people leave the treatment group than the control group over time.

Attrition is an important factor in assessing the value of any research. It is not a matter which should be hidden in the report of the research.