Study Guides (248,414)
Canada (121,549)
Psychology (623)
PSYCH 339 (4)

PSYCH339 NOTES final.docx

13 Pages
Unlock Document

James Beck

March 21 Selection Decision Making continued 1. Top-down 2. Bottom-up - trying to screen people out, want to screen out who can’t even fake, or didn’t mention keywords, good for multiple hurdle approach (make it less expensive) 3. Multiple Hurdle - pass one to be able to take second - multiple predictors - test that’s simplest and cheapest to go first 4. Multiple Cut-Off - also multiple tests but everyone takes all but minimum cut scores set for each test - if anyone doesn’t pass the cut-off for any of the test, screen out - if one or all of tests are absolutely critical for the job - e.g. surgeon, pilot, astronaut 5. Compensatory - high scores one one predictor can compensate for low scores on another - breaks down into clinical and mechanical - clinical: assessor uses judgement and expertise; appealing because people think they are good at it and look at broken legs Problem: people are bad at it, different definitions of broken legs, unreliable measurement of broken legs - mechanical: combining info in some systematic way; unit weighting simple add up or optimally weight factors with multiple-regression. Each person gets weight applied consistently, predicted performance is derived systematically Multiple regression chooses the weights that maximizes R 6. Banding - all people within the same “band” are considered equivalent - band defined by SEM - can’t differentiate people within same confidence interval, no preference - not using strict top-down - if need 1 but band has 6 in the band, then throw in something else like conscientiousness E.g. Standard dev = 1 rxx = 0.75 SEM = standard dev square root (1-rxx) = 0.5 Band = ~ 2 * 0.5 = 1 Start with z-score of first on the list. Next band starts after everyone in first band Legal and Ethical Issues in Hiring Protected Classes in ON - race/colour - religion - physical or mental disability - dependence on drugs/alcohol - age/sex/sexual orientation - national or ethnic origion/language - pregnancy - marital status/family status Means: cannot hire based on standing on above variables Nor other HR functions - recruitment - promotion - termination Direct discrimination: - an illegal conscious act or practice to discriminate on a prohibited ground - overt, explicit Indirect discrimination: - an act that does not directly discriminate but results in excluding or restricting a group of persons who are identified by a prohibited ground e.g. test of hiring results, no women? Synonyms: - Adverse impact (U.S.) - constructive discrimination - adverse effect 4/5 Rule - selection exhibits adverse impact if: - selection ratio for protected group is less than 80% selection ratio of majority group (different issues being argues about in legal arena but this is basis) Example: A) Company needs to hire 100 people, 200 applied Minority hires: 50 Minority applicants: 100 Minority selection ratio: 0.5 Majority hires: 50 Majority applicants:100 Majority selection ratio: 0.5 Adverse Impact ratio = Minority selection ratio/majority selection ratio = 1 No Adverse Impact B) Minority selection ratio: 0.4 majority selection ratio: 0.5 Adverse Impact Ratio = 0.4/0.5 = 0.8 No Adverse Impact C) Minority selection ratio: 0.4 majority selection ratio: 0.55 Adverse Impact Ratio = 0.4/0.55 = 0.73 If adverse impact, employer has to show: - trait being selected for is a B.F.O.R.  Bona fide occupational requirement (hiring for actress, have to be women)  reasonable accommodations can’t be made - selection system is valid, critical point; system must predict performance (not designed to screen out a group of people) - no equally valid alternatives with less adverse impact available (include other factors that has less adverse impact) Non-cognitive factors reduce adverse impact so include other factors; cognitive factors cause more adverse impact. Affects multiple hurdle results, cognitive factor test done first, more AI. How does this happen? Subgroup differences - Cognitive tests  Race - Physical tests  Sex Cognitive tests - group-level - can’t tell us about an individual - more variance within groups than between but over so many people, get group “trend” Not a measurement issue; cognitive tests aren’t “biased” Differences in background (Experiences, knowledge) Stereotype threat (self-fulfilling prophecy) Slope differences: This would be bias – not what we find, would be a big problem big scale - minority slope shallower We find intercept differences, differences in average score - lower performing group gets over-predicted, vice versa - does lead to adverse impact Even though test isn’t biased - still group-mean differences - still adverse impact Bias: technical issue Fairness: social issue Implications for: - withdrawal from selection system - psychological well-being off applicant - litigation Early work: focused on construct/method; how applicants feel about the method Unfortunately, how well it works doesn’t always match favourability. More recent work: focus on justice perceptions (do you think it’s fair) Applicants expect: transparency, objectivity, feedback, job information, participation, humane treatment Correlations with procedural justice perceptions: outcome favourability 0.24 one of lowest; perceived predictive validity 0.63, opportunity to perform 0.56, job relatedness0.61 highest Withdrawal from selection procedures - if people dropped out of selection system, increases selection ratio and decrease utility; has a drastic impact. Determines if they come back for the second hurdle, would be very bad if people perceived unfairness and dropped out Only 5% didn’t show up for the second hurdle and said it. If got explanation, more likely to reapply, significant but very small So applicant reactions don’t have huge effect on withdrawal, but they do have some effect. People experience less affective (emotional) distress when procedural justice perceptions are high. Less likely to file a lawsuit against hiring organization, why take the risk? March 26 Training - formal procedures - facilitates learning - increase behaviour beneficial to org Personnel selection selects for Abilities, Personality/Interests, Knowledge and Skills; K and S are also under training and development 3 par
More Less

Related notes for PSYCH 339

Log In


Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.