I/O Psychology Flashcards
3 problems that can result in discrimination
- adverse impact
- unfairness
- differential validity
adverse impact
80% or 4/5ths rule: % of minorities selected must be at least 80% of the non-minorities selected
multiply the selection rate for non-minorities by 0.8 to assess
unfairness
when minorities and non-minorities score differently on a predictor test (e.g., MCAT) but perform similarly on the criterion (e.g., med school)
differential validity
significantly different criterion-related validity coefs. for different groups on the same test
ex. the test is more valid for predicting men’s performance than women’s performance
rarely occurs - previous lit. on this largely due to low sample size
job analysis
-describes the nature of the tasks of a job
-gives info re: the job description (tasks), job specifications (requirements), and necessary data for selection procedures, determining wages, training needs, etc.
-data obtained through interviews and critical incident technique
job description
describes tasks of the job
job specifications
requirements for the job
job evaluation
determining wages/worth of the job
biodata
biographical info, an option for job selection
includes:
1) the standard application
2) the weighted application blank (assigns weights to certain variables on the standard app)
3) biographical inventory (aka BIB)
biographical inventory (aka BIB biographical information blank)
covers applicant’s life in more detail, Qs correlated w/ desirable/undesirable work bxs, validated against specific criterion
-very good predictor of job success
-good predictor of turnover
-time-consuming to develop, potentially costly
tests of cognitive ability and aptitude
-good predictors of job success
-may result in discrimination
Griggs vs. Duke Power Co.
issue of testing in the workplace
ruled that tests measuring broad abilities had racial bias embedded and were unfair to used for hiring and promotion decisions
established that tests measure skills necessary for a particular job
interest tests (e.g., Holland’s self-directed search)
-correlate w/ job satisfaction
-poor prediction of job success
work samples
-high content and criterion validity
-tend to be more valid for minorities compared to other types of testing
test batteries
used w/ upper management
good predictor of job performance
assessment centers/situational testing
simulated job situation, very good criterion-related validity
1) in-basket technique: presented w/ problems/Qs managers would expect to find when they return from vacation
2) leaderless group discussion
multiple regression approach to selection
-compensatory
-low scores on one predictor (e.g., low GPA) can be compensated for by high scores on another (e.g., high SAT, good interview)
-involves computing multiple regression equation
multiple cutoff
-noncompensatory
-only applicants who meet/exceed the cutoff on EACH of the predictors will be considered
multiple hurdle
-noncompensatory
-predictors applied in a particular order, must pass the cutoff on the first in order to continue in selection process
-efficient
BARS (behaviorally anchored rating scales)
-subjective method of performance appraisal
-bx anchors based on critical incidents, and employees are rated on these anchors (e.g., bxs associated w/ successful job performance)
-scale development can be expensive and time-consuming