The term refers to the tendency of raters – managers, supervisors or interviewers – to evaluate and grade a majority of employees or subjects as average or away from the extreme points of the scale. As a result, outliers are difficult to recognise. The impact of the central tendency error increases with increase in rating scale size – but shorter rating scales are also less exact.
The term ‘central tendency’ is rooted in statistics where it is used to refer to the average or middle range of a distribution. Such errors usually occur because of a natural tendency to provide “middle-of-the-road” ratings (such as teachers or managers neither wishing to claim someone has attained mastery or perfection nor willing to assign extremely negative scores) or because of habituation. It may be also be due to items or questions that tend to cause people to choose less extreme responses.
Therefore, scales and surveys must be designed such that there is good variation in the responses. Forcing comparable ratings, that is to rank relatively rather than absolutely, also helps distinctions be brought out.