PSYC 245 - Staffing Decisions

Staffing decisions::Decisions tied to recruiting, selecting, promoting, and terminating employees. Purpose is to hire and keep people who will be successful on the job.

Recruitment::Attracting the right people to apply. Based on job analysis; uses job postings, referrals, and recruiters at job fairs.

Selection::Collecting and evaluating information about an individual in order to extend an offer of employment. Hiring the right people.

Promotion::Putting the right people in power.

Deselection::Getting rid of the wrong people (termination/layoff).

Job analysis::Identifies what KSAOs we are looking for. Job postings should be based on it.

Selection assessment::How we assess KSAOs. Gives us scores.

Four questions of selection::What KSAOs are we looking for (job analysis)? How do we assess them (selection assessment)? How do we make selection decisions? How do we evaluate the outcomes?

Clinical (intuitive) method::Combining information based on human judgment, such as an expert thinking about all the information. Weighting of factors is informal. Unreliable and may lead to biased decisions.

Statistical (actuarial) method::Using a mathematical formula to evaluate and/or combine information. More reliable, less biased, and produces superior decisions.

Total/average scores::Adding up or averaging scores across selection measures, then using a cut-off score or rank order.

Unit-weighting::Total/average score approach where each variable is assumed to be equally important in predicting performance.

Weighted average::Combining scores with different weights; can change the rank order of candidates compared to a simple average.

Multiple regression::Statistical method that weights all predictors at once and shows how much "pull" each variable has on the criterion (e.g., job performance).

Predictor::A variable used to predict the criterion (e.g., intelligence, personality).

Criterion::The outcome being predicted (e.g., job performance).

Hurdle system::Sets cut scores for each selection method to screen out candidates sequentially. Used to narrow a large applicant pool; more expensive hurdles (e.g., interviews) are placed later.

Score banding::Grouping candidates with similar scores into a category so everyone in the range is treated as the same. Removes meaningless rank-order differences caused by measurement error. Can sometimes increase diversity.

Subgroup norming::Developing separate ranked lists for people within a certain demographic, so each group is compared only to its own members. Intended to increase diversity but often backfires. Illegal under the Civil Rights Act of 1991.

Compensatory selection system::Candidates can make up for a weakness in one attribute with strength in another. Example: multiple regression, where requirements are considered simultaneously.

Noncompensatory selection system::Weaknesses cannot be overcome. Example: hurdle system, where requirements are considered stepwise (a legally blind applicant is not eligible to be a bus driver).

Validity::The extent to which we are hiring the correct people. How well we can predict job performance from selection scores.

Criterion-related validity::Establishing that a selection system works by correlating selection scores with job performance scores.

Correlation (r) in validity::r = 1.0 is perfect prediction (not possible); r = 0 means the predictor is useless; 0 < r < 1.0 is what is actually observed.

Selection ratio (SR)::The proportion of applicants actually selected into positions. SR = number of hires / number of applicants (100 applicants for 10 positions = .10).

Base rate (BR)::The proportion of employees considered successful performers (40 of 100 satisfactory = .40).

Cut score::A predictor score below which candidates will not be hired. Raising the cut score makes the selection ratio smaller.

Criterion-referenced cut score::Cut score that corresponds to a desired level of performance (uses multiple regression to find the minimum predictor score for proficient performance).

Norm-referenced cut score::Cut score based on how test takers score compared to other test takers; does not formally consider performance.

True positive::Selecting someone who performs acceptably well (correct decision).

True negative::Not selecting someone who would not have performed well (correct decision).

False positive::Selecting someone who does not perform acceptably well (incorrect decision).

False negative::Failing to hire someone who would have performed acceptably well (incorrect decision).

When base rate increases::More true positives and false negatives; fewer true negatives and false positives.

When selection ratio increases::More true positives and false positives; fewer true negatives and false negatives.

Utility::The economic return on investment of HR interventions like staffing or training. Addresses the cost/benefit ratio of one staffing strategy versus another.

Utility considerations::Cost of selection procedures per candidate; number of candidates vs. positions; effect of lower performance overall; how well current procedures work.

Fairness::A subjective judgment, so what counts as a fair selection process is largely defined by law and professional guidelines.

Adverse treatment::Employer knowingly and willfully treats minority applicants differently than majority applicants. Intentional and direct discrimination.

Adverse impact::Employment practice incidentally treats minority applicants differently than majority applicants. Unintentional, but still creates unequal outcomes.

Adverse impact ratio::Minority group's selection ratio divided by majority group's selection ratio.

80 percent (4/5ths) rule::If the protected group receives less than 80% of the desirable outcomes the majority group receives, there is evidence of possible adverse impact. Crude guideline affected by sample size; courts decide based on the whole case.

Adverse impact example::10 of 20 men promoted (50%), 3 of 10 women promoted (30%). 30/50 = 60%, which is under 80%, so there is evidence of adverse impact.

Burden shift in adverse impact cases::After adverse impact is shown, the employer must demonstrate the practice was job related and therefore valid. I/O psychologists often serve as expert witnesses.

Predictor weight combinations::Strategy to reduce adverse impact by combining predictors (e.g., cognitive ability and conscientiousness).

Diversity/minority recruiting::Strategy to reduce adverse impact by increasing the minority applicant pool. Recruiting on demographics alone is not enough; need to improve quality and quantity.

Affirmative action::Programs that seek to rectify the effects of past employment discrimination by increasing representation of certain groups (women, minorities, individuals with disabilities).

Weak preferential treatment::Affirmative action approach that uses group membership as a tiebreaker.

Strong preferential treatment::Affirmative action approach that uses differential criteria to meet diversity goals.

Civil Rights Act of 1964 (Title VII)::Required selection processes to be job-related, made adverse impact a standard, and created the EEOC to enforce the law.

EEOC::Equal Employment Opportunity Commission; created by the Civil Rights Act of 1964 to enforce it.

Uniform Guidelines on Employee Selection Procedures::Selection guidelines published by the EEOC in 1978.

Bona Fide Occupational Qualification (BFOQ)::Allows gender, religion, or race-based decisions when there is a genuine business necessity. Examples: towel attendant in a ladies restroom, male-only Catholic priests, mandatory retirement ages for pilots and bus drivers.

Civil Rights Act of 1991::Made race- or gender-norming illegal: no separate cutoffs, no separate norms, no bonus points for race/gender/ethnic membership.

Equal Pay Act (1963)::Men and women doing equal work in an organization must be paid the same. "Equal" is defined by skill, effort, responsibility, and work conditions.

Age Discrimination in Employment Act (ADEA; 1967)::Prohibits discrimination against anyone over age 40. Provisions match Title VII. Protects older workers, who are often the highest paid.

Americans with Disabilities Act (ADA; 1990)::Prohibits discrimination against those with disabilities.

ADA definition of disability::A physical or mental impairment that substantially limits one or more major life activities (e.g., blindness), or a record of such an impairment (e.g., cancer survivors). Obesity and drug and alcohol abuse are not covered.

Family Medical Leave Act (1993)::Federal law listed for pregnancy as a protected class.

Protected classes and federal law (U.S.)::Age (over 40): ADEA 1967. Disability: ADA 1990. Race, national origin, religion: Civil Rights Act 1964/1991. Sex: Civil Rights Act 1964/1991 and Equal Pay Act 1963. Pregnancy: Family Medical Leave Act 1993.