Angus Okechukwu Unegbu, James J. Adefila
The main focus of this paper is on Canonical Correlation Analysis as a bias scoring detector, using American University of Nigeria as a study case. In carrying out the research, we stipulated three null hypotheses. These hypotheses were stipulated after envisaging a problem that American University of Nigeria in a bid to keep with her value statement. The value statement stipulates that in all her activities, she will demonstrate the highest standards of integrity, transparency and academic honest but the University is faced with dwindling funding resulting in a cross road as to how to appraise and select candidates for promotion from numerous non-queuing applicants who have to secure highest weighted recommendations from various levels of promotion committees. The need for a bias free selection technique becomes very essential for adaption in analyzing applicants’ scores from the various promotion committees. Scores of Candidates used in the research were obtained via secondary sources. Manual computations were first used to process the obtained scores and Computer SPSS Canonical Correlation Analysis were later used to process the results in order to test stated hypotheses. We found that Canonical Correlation Analysis has the capacity to detect bias scoring, overbearing score-weight influence and has the ability to discriminate between promotable and non-promotable candidates. We recommend its use to American University of Nigeria and other Universities of excellence. We also recommend for more research into the relevance of the statistical tool in the areas of Accounting and Business decision endeavors
Share this article