Open AccessApplication of Rasch to Verify Different Building Competency Sub Constructs for Competency Evaluation of Graduates in Nigeria
Author(s) -
Shirka Kassam Jwasshaka,
Nor Fadila Mohd Amin
Publication year - 2020
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.f1396.089620
Subject(s) - rasch model , content validity , respondent , psychology , reliability (semiconductor) , construct validity , medical education , face validity , applied psychology , construct (python library) , consistency (knowledge bases) , ibm , statistics , computer science , mathematics , psychometrics , medicine , artificial intelligence , political science , power (physics) , physics , quantum mechanics , law , programming language , materials science , nanotechnology
Incessant concerns from employers and private sectors about the incompetence of graduates in Nigeria call for the creation of an assessment tool that could verify their skills. But there is no clear generally accepted and validated assessment instrument available for evaluating graduate performance. The aim of this study was to develop a valid instrument for assessing competency levels of building construction graduates in Nigeria. Survey design was adopted to obtain expert opinions on the validity of the sub-constructs and the related items about the employers 'needs. Three experts from the academic, public, and private sectors subjected the survey instrument to face, content, and construct validity and reliability. The survey instrument, which was analysed using IBM SPSS and WINSTEP version 3.73.3 was answered by a total of 200 building experts selected by proportionate stratified sampling technique. The consistency of the instrument was determined by fit statistics and point measure correlation (PTMEA Corr), for construct validity. The results revealed a very good items and person reliability of 0.97 and 0.94 respectively. Likewise, appropriate PTMEA Corr range from 0.36 to 0.68. Infit and outfit means square range obtained between 0.58 to 1.39. The findings give students, employers and academic institution a realistic and theoretical interpretation of the reality of labour market needs.