Multi-temporal analysis of regression diagnostics for students' performance in property development and co-requisite courses

Abstract

Property development and other co-requisite courses have been embedded within academic programs in Estate Management. Among the objectives of this study include identifying the co-requisite courses that accounted for a variation in students' academic performance in property development, and to examine the characteristics of the diagnostics of ordinary least square (OLS) regression model for this relationship over time. Data for this study comprised 2007/2008 to 2014/2015 (8-year) scores for the 1st- and 2nd semester property development and its co-requisite courses in the Department of Estate Management of the Federal Polytechnic Idah, Nigeria. Tools of analysis comprised descriptive statistics (mean score, standard deviation, and coefficient of variation), fitted trend equations, multiple linear regression analysis, and a multi-temporal visualization of the regression diagnostics comprising *ZRESID vs. *ZPRED, normality tests for residuals, and the Durbin-Watson test. The variation in the scores for property development courses during the 8-year period was attributed to specific 1st- and 2nd semester co-requisite courses. EST311 (Valuation I) was found to be the most consistent regressor for EST313 (Property Development I) in the 1st semester, while EST325 (Estate Accounting), was the most consistent regressor for EST323 (Property Development II) in the 2nd semester. Multi-temporal diagnostics analyses did not provide any sufficient evidence to invalidate the OLS regression models for the 1st- and 2nd semester Property Development courses and their co-requisites. Stakeholders' commitment to the teaching and learning of the co-requisite courses was recommended as triggers for a sustained pass rate in the property development courses.

Keywords

• Property development
• co-requisite courses
• regression
• multi-temporal analysis
• students' performance