Return to search

Demographic similarity effects in assessment centre ratings : an investigation of assessor bias in the South African context

Thesis (MComm)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: Assessor bias remains an interesting and important topic for Assessment Centre
(AC) practitioners, candidates and clients, but it has received relatively little
research attention. By using social identity theory, this research project studied the
effects of demographic characteristics (gender, age and race) on ratings in
assessment centres. We studied main and interaction effects of assessor and
assessee demographic variables on ratings in a national sample of managers (N =
220). Descriptive statistics (means, SD) and inferential statistical analysis
(moderated multiple regression) were used to test for both main and interaction
effects (between assessor and assessee characteristics) of demographic variables on
ratings. The regression analysis revealed a highly statistically significant (p < .001.)
gender main effect (for assessor) — male assessors consistently rated assessees
higher that female assessors did. Neither assessor nor assessee race had a main
effect on ratings (p > .05). Age showed no main or interaction effects, although
evidence of possible interaction effects was seen in descriptive statistics.
Demographic similarity did not seem to affect assessment centre ratings in the
present study. Recommendations for future research and AC practice are made. / AFRIKAANSE OPSOMMING: Assessorsydigheid het tot dusver betreklik min navorsingsaandag ontvang,
alhoewel dit steeds 'n interessante en belangrike onderwerp vir alle
takseersentrumbelanghebbendes verteenwoordig. Vanuit ‘n sosiale
identiteitsteorieraamwerk, het hierdie navorsingsprojek die invloed van
demografiese kenmerke (geslag, ouderdom en ras) op takseersentrumbeoordelings
ondersoek. Beide die hoofeffekte en interaksie-effekte van assessors en assessees
se demografiese veranderlikes op beoordelings is ondersoek. Die studie het ‘n
nasionale steekproef van bestuurders (N = 220) gebruik om die hipoteses te toets.
Beskrywende statistiek (gemiddelde en standaardafwykings) en inferensiële
analises (gemodereerde meervoudige regressie-analise) is gebruik om beide hoofen
interaksie-effekte (tussen die assessor en assessee se demografiese
veranderlikes) op beoordelings te toets. Die regressie-analise het hoogs statisties
betekenisvolle (p < .001) geslagshoofeffekte (assessor) aan die lig gebring -
manlike assessors het konsekwent hoër tellings as vroulike assessors aan assessees
toegeken het. Rasgroep (van beide assessor en assessee) het geen hoofeffek op
beoordelingstellings gehad nie (p > .05). Ouderdom het nie hoof- of interaksieeffekte
getoon nie, alhoewel bewyse van moontlike interaksie-effekte bespeur is in
beskrywende statistiek. Die resultate van die studie toon dat demografiese
gelyksoortigheid tussen die assessee en assessor nie ‘n beduidende effek op
takseersentrumbeoordelings het nie. Aanbevelings vir toekomstige navorsing en
takseersentrumpraktyk word gemaak.

Identiferoai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:sun/oai:scholar.sun.ac.za:10019.1/20411
Date03 1900
CreatorsDe Villiers, Christine Marguerite
ContributorsDe Kock, Francois, Stellenbosch University. Faculty of Economic and Management Sciences. Dept. of Industrial Psychology.
PublisherStellenbosch : Stellenbosch University
Source SetsSouth African National ETD Portal
Languageen_ZA
Detected LanguageUnknown
TypeThesis
Format124 p.
RightsStellenbosch University

Page generated in 0.0016 seconds