IOP4863
Assessment 05
Unique No: 148704
Due Date 6 October 2025
, IOP4863
Assessment 05
Unique No: 148704
Due Date: 6 October 2025
Personnel and Career Psychology
PART A: PERSONNEL PSYCHOLOGY
Question 1.1: What are differential validity and differential prediction, and how do
they apply to the selection committee’s task?
Differential validity refers to whether a selection tool (such as a cognitive ability test) is
equally valid for predicting job performance across different demographic groups. If the
test predicts accurately for one group but not another, differential validity is present,
which makes its use problematic (Schmidt & Hunter, 1998).
Differential prediction occurs when test scores do not predict performance equally
across groups. For example, the same score on a cognitive test might overestimate the
job performance of one group while underestimating another. This introduces
systematic bias (Cascio & Aguinis, 2019).
In the committee’s task, both concepts highlight the importance of using tools that have
been validated across groups in South Africa. If a test shows differences in prediction or
validity, the committee must either adjust scoring, combine it with other methods (such
as structured interviews or work samples), or avoid its use altogether. This ensures
compliance with the Employment Equity Act and prevents unfair discrimination.
Assessment 05
Unique No: 148704
Due Date 6 October 2025
, IOP4863
Assessment 05
Unique No: 148704
Due Date: 6 October 2025
Personnel and Career Psychology
PART A: PERSONNEL PSYCHOLOGY
Question 1.1: What are differential validity and differential prediction, and how do
they apply to the selection committee’s task?
Differential validity refers to whether a selection tool (such as a cognitive ability test) is
equally valid for predicting job performance across different demographic groups. If the
test predicts accurately for one group but not another, differential validity is present,
which makes its use problematic (Schmidt & Hunter, 1998).
Differential prediction occurs when test scores do not predict performance equally
across groups. For example, the same score on a cognitive test might overestimate the
job performance of one group while underestimating another. This introduces
systematic bias (Cascio & Aguinis, 2019).
In the committee’s task, both concepts highlight the importance of using tools that have
been validated across groups in South Africa. If a test shows differences in prediction or
validity, the committee must either adjust scoring, combine it with other methods (such
as structured interviews or work samples), or avoid its use altogether. This ensures
compliance with the Employment Equity Act and prevents unfair discrimination.
