IOP4863
Assignment 5
Unique No:148704
DUE: 6 October 2025
,IOP4863 ASSESSMENT 05 - DETAILED ANSWERS
PART A: PERSONNEL PSYCHOLOGY
QUESTION 1.1: Differential Validity and Differential Prediction (10 marks)
Differential Validity:
Differential validity occurs when a selection instrument (test, interview, assessment) has
different validity coefficients for different subgroups of the population. This means the
test predicts job performance more accurately for one group than another.
Practical Application to Selection Committee:
• As chair, I must ensure that any selection method chosen demonstrates
comparable validity across demographic groups (race, gender, age)
• Example: If a cognitive ability test predicts managerial success with r=0.65 for
white candidates but only r=0.35 for black candidates, this indicates differential
validity
• The committee must conduct validation studies separately for different subgroups
before implementing any assessment
• If differential validity is found, we must either modify the instrument or select
alternative methods that show consistent validity across groups
Differential Prediction:
, Differential prediction occurs when the regression lines predicting job performance from
test scores differ systematically between groups. This means the same test score
predicts different levels of job performance for different groups.
Practical Application:
• The committee must analyze whether test scores systematically over-predict or
under-predict performance for certain demographic groups
• Example: If a score of 75 predicts "good" performance for Group A but "excellent"
performance for Group B, using a single cutoff score would be unfair
• We must examine intercepts and slopes of regression equations for each
subgroup
• If differential prediction exists, separate prediction equations or adjusted cutoff
scores may be necessary
• This is critical for legal defensibility under the Employment Equity Act
Committee Implementation:
• Conduct concurrent or predictive validation studies
• Calculate separate validity coefficients and regression equations for protected
groups
• Document all findings thoroughly for legal purposes
• Make evidence-based decisions about test utility and fairness
QUESTION 1.2: The Concept of Fairness in Employment Decisions (5 marks)
Definition and Application:
Fairness in employment selection refers to ensuring that all candidates have equal
opportunity to demonstrate their capabilities and that selection decisions are based
solely on job-relevant criteria without bias or discrimination.
Key Dimensions of Fairness:
Assignment 5
Unique No:148704
DUE: 6 October 2025
,IOP4863 ASSESSMENT 05 - DETAILED ANSWERS
PART A: PERSONNEL PSYCHOLOGY
QUESTION 1.1: Differential Validity and Differential Prediction (10 marks)
Differential Validity:
Differential validity occurs when a selection instrument (test, interview, assessment) has
different validity coefficients for different subgroups of the population. This means the
test predicts job performance more accurately for one group than another.
Practical Application to Selection Committee:
• As chair, I must ensure that any selection method chosen demonstrates
comparable validity across demographic groups (race, gender, age)
• Example: If a cognitive ability test predicts managerial success with r=0.65 for
white candidates but only r=0.35 for black candidates, this indicates differential
validity
• The committee must conduct validation studies separately for different subgroups
before implementing any assessment
• If differential validity is found, we must either modify the instrument or select
alternative methods that show consistent validity across groups
Differential Prediction:
, Differential prediction occurs when the regression lines predicting job performance from
test scores differ systematically between groups. This means the same test score
predicts different levels of job performance for different groups.
Practical Application:
• The committee must analyze whether test scores systematically over-predict or
under-predict performance for certain demographic groups
• Example: If a score of 75 predicts "good" performance for Group A but "excellent"
performance for Group B, using a single cutoff score would be unfair
• We must examine intercepts and slopes of regression equations for each
subgroup
• If differential prediction exists, separate prediction equations or adjusted cutoff
scores may be necessary
• This is critical for legal defensibility under the Employment Equity Act
Committee Implementation:
• Conduct concurrent or predictive validation studies
• Calculate separate validity coefficients and regression equations for protected
groups
• Document all findings thoroughly for legal purposes
• Make evidence-based decisions about test utility and fairness
QUESTION 1.2: The Concept of Fairness in Employment Decisions (5 marks)
Definition and Application:
Fairness in employment selection refers to ensuring that all candidates have equal
opportunity to demonstrate their capabilities and that selection decisions are based
solely on job-relevant criteria without bias or discrimination.
Key Dimensions of Fairness:
