3 Selection: Analysing jobs, competencies and selection methods Flashcards
What is best practice selection?
Hvorfor gjøres det ofte ikke optimalt?
Krever:
- jobbanalyse
- tilpassede selection tools based on jobanalysis
- feedback fra validationstudies (vanskelig, tidskrevende, dyrt)
Ulike datatyper for jobbanalyse
- spørreskjema
- intervju
- observajson
- task-analyse
- work records
-Analysis of existing policy documentation
Hva er Competency models?
- Competency models define the key behaviours associated with performance in the target job role.
- While generic competency frameworks may be a tempting and cost-effective option, bespoke is best as every job role.
- dynamisk
What is Flanagan’s (1954) procedure known as the CIT
“Critical incident technique, CIT”
Be intervjuobjekt eller gruppe beskrive en spesifikk jobb-atferd i forkant/etter kritisk hendelse (bra/dårlig).
Gir jobbanalysør mye info om jobb og atferd
Ulempe: fokus på ekstremsituasjoner og ikke rutinejobb
How is reliability relevant?
-reliabilitet til selection instrument ekstremt viktig
-i hvilke grad måler utvelgelsesmetoden konsistent under ulike situasjoner
External reliability
-Test-retest reliability
(tester en gruppe på ulike tidspunkter, forventes likt resultat ish)
-Parallel forms of reliability
(to ulike tester som måler samme konstrukt, forventes å ha like resultater hvis reliabel)
Internal reliability
-Cronbach’s coefficient alpha
(statistisk analyse som tester om data er internt konsistente. Alpha over 0.7 = god intern concistency)
-Split-half method
(data delt i to og undersøkes om skårer er nogenlunde like)
-Kuder-Richardson 20
(spess form for Cronbachs Alpha for dikotone data/spørsmål)
The ideal validation of a selection processes. What is it?
Criterion related validity
- Criterion-related validity is the strength of the relationship (correlation) between the predictor and the criterion/work performance (Task performance, citizenship performance and counterproductive work behaviour).
- validitetskoeffisient = korrelasjon mellom prediktor og kriterion
To measure criterion related validity: Predictive validity designs (follow-up design) are more rigorous than _________ _________ _______, but their implementation usually presents several practical problems.
concurrent validity designs
- Jobbanalyse finner kriterier, tester allerede ansatte istede for å teste de man undersøkte
Different types of validity in selection (8 stk)
- Faith validity
- Face validity
- Content validity
- Incremental validity
- Construct validity
- Political validity
- Criterion validity: Concurrent
- Criterion validity: Predictive
To be effective, selection methods must be both_____ and_____.
reliable, valid
3 source of error when estimating
the validity of selection procedures
- Sampling error
(ofte lite sample 50-150, uteliggere) - Poor measurement precision
(true score plus error, mål lite reliabelt? isåfal minskes
korrelasjon mellom prediktor og kriterion) - Restricted range of scores(urepresentativt)
Sample used in validation is a selected sample, generaliseres til hele søkergruppe (unselected sample), kan få lavere korrelasjon enn det egentlig er mellom prediktor og kriterion
Using selection procedures with good predictive validity is always important but, unfortunately, procedures with good predictive validity alone do not guarantee that a selection procedure will be cost-effective, aften refered to as _____ ______.
2 important factors determining cost-effectiveness?
Financial utility
- the selection ratio, i.e. number of jobs/number of candidates
- the financial benefit of improved job performance.
In most situations it is best to use a _________ __ ________selection techniques to ensure fairness, accuracy and bolster validity.
combination of several
Cognitive ability tests remain consistent in producing ____ criterion-related validities for a wide range of jobs.
good
Tankekart
