I/O Flashcards
Job analysis vs job eVALUation
Analysis –> describes job REQUIREMENTS
* job-oriented methods = info re: tasks
*worker-oriented methods = info re: worker (“KSAOs”)
Evaluation –> establishes COMPARABLE WORTH thru compensable factors (skills, education, degree of responsibility, consequences of errors)
KSAOs
Knowledge
Skills
Abilities
Other characteristics
Types of criterion mesures
Objective: quantitative; often inadequate, may be biased due to situational factors; often unavailable to managerial/admin/prof jobs
Subjective: most frequently used; reflect judge of rater
Subjective raters
self: most lenient, less halo effect
supervisor: most reliable
peer: predicts training success and promotions
subordinate, peer and supervisor: agree more with each other than with self
characteristics of criterion measures
ultimate - vs- actual criterion
relevance –> construct validity
deficiency –> degree actual criterion DOES NOT measure all aspects
contamination
subjective criterion measures can be ____ or ___
relative or absolute
relative vs absolute criterion measures
relative--> compares 2 or more employees helps reduce rater biases forces rater to rate high or low prohibited for most federal jobs ** raters/ratees dislike them low employee feedback on performance
absolute–> only rates one employee’s performance
Types of relative subjective criterion measures
Paired comparison forced distribution ("grading on the curve")
Types of absolute subjective criterion measures
critical incident technique
forced choice
graphic rating scale
BARS
predictor vs criterion
predictor = selection criterion = ON the job
reliability vs validity
reliability = unaffected by measurement error validity = measures what its designed to assess
what type of validity is most important in IO?
criterion-related validity
steps of criterion-related validity
1) job analysis
2) select/develop predictor
3) administer predictor (to applicants) and criterion (to current employees)
4) correlate predictor and criterion scores (criterion-related validity coefficient)
5) check for adverse impact ***
6) evaluate incremental validity
7) cross-validate (redo steps 3-6 w/new sample)
80% Rule
majority group hire rate X .80 = minimum hire rate of minority
what is shrinkage?
when cross-validation coefficient is smaller because chance factors from 1st sample are not present
Causes of adverse impact
differential validity (valid for one group, not other) unfairness (diff groups get diff scores on predictor, but similar scores on criterion)
selection ratio
job openings: applicants
LOW ratio preferred (one job, many applicants)
Can RAISE predictor cutoff to reduce hiring a false +
(raise predictor, lower criterion)
Base rate
% of employees performing satisfactory without using proposed predictor (from 0 - 1.0)
best base rate for incremental validity?
Moderate: (close to .5)
Taylor-Russell tables
Best:
- -low validity coefficient (.30)
- -low selection ration (1:100)
- -Moderate base rate (.45)
multiple predictors (3)
Multiple regression–>”compensatory” - exceptional performance in one area can compensate for poor areas
multiple cutoff–> need minimum score on each predictor
multiple hurdles–>need min score in order to advance to next predictor (saves time and $)
Common predictors of job performance (8)
1) general ability (cognitive) tests
2) job knowledge tests
3) personality tests
4) interest tests
5) biota
6) interviews
7) work (job) samples
8) assessment centers
“big 5” personality
Conscientiousness ** (best predictor) neuroticism extraversion openness agreeableness
common predictors with highest validity coefficients are ____ and ____.
general ability tests (highest)
job knowledge tests (job-specific)