The concept of unfair labour practice has been introduced into South Africa through the Labour Relations Act and it is argued that certain psychometric testing practices can be interpreted as falling under the definition of an unfair labour practice. Empirical results are presented indicating that different ethnic groups obtain significantly different mean ability test scores. A case-study is cited to show that this would result in underprediction of performance on a criterion for the lower scoring of two groups if test scores are regarded as comparable. It is argued that any interpretation of psychometric data in South Africa that does not take account of possible differences between ethnic groups is likely to ignore a significant moderator variable, given the history of ethnically based discriminatory practices in this country. Arguments based on meta-analytical research in the USA, to the effect that psychometric ability tests do not discriminate unfairly against disadvantaged groups if the same tests and norms are used, should not be assumed to hold in South Africa. Various conceptions of what constitutes fairness in selection are considered, and it is concluded that there is a need for employers to make explicit in their selection policies the trade-off between economic and social costs of employment practices. Finally, some implications for users of psychometric tests in industry are considered, in order to forewarn of likely developments in this field.
The contribution of the nurse manager to healthy labour relations: Part 2
