ASSESSMENT OF THE AWARENESS OF DECISION SUPPORT SYSTEM TO RESIDENTIAL PROPERTY VALUE DETERMINATION IN LAGOS, NIGERIA.

Abstract

The ability of Estate Surveyors and Valuers to accurately determine the value of real estate plays a strategic role in an economy and influences the investor’s decisions. However, consequent upon the rising cases of valuation inaccuracies and variances which has been largely blamed upon the limitations associated with the adoption of traditional methods of valuation. The current research examines the awareness of Decision Support Systems to real estate valuation with a view to determining their respective performances in residential property valuation in Lagos Metropolitan property market, Nigeria. A total of 3,079 datasets of transaction sales prices of different residential property types were sourced from the database of one hundred and ninety-four (194) practicing firms in the Lagos metropolitan property market using the combination of online and structured questionnaires. Descriptive Statistics, Correlation Analysis, Chi-Square, and Factor Analysis were used to analyze the objectives of the research. Six (6) advanced computer-based valuation techniques comprising Hedonic Model, Artificial Neural Network, Random Forest, CatBoost, XGBoost and LGBM were used in predicting the market value of the properties in the study area. A total of 80% of the 3,079 datasets collected representing 2,463 datasets were set aside for model building and training while the remaining 20% representing 616 datasets were also set aside to test the predictive ability of the six (6) developed models. Four model performance metrics such as Coefficient of Determination (R2), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) were adopted to determine the respective performances of the selected models. The results of the Weighted Mean Score, Correlation, and Chi-Square reveal amongst others that there is no significant relationship between Valuer’s educational & professional qualifications and awareness of Decision Support Systems indicating a low level of awareness of these techniques on the part of the Valuers. The results of the Factor Analysis with a KMO of 0.789 which generated a cumulative percentage of 73.713% based on an eigen level of 5.653 also revealed that the absence of centralized databank, lack of market data & inadequate technical know-how of the techniques are the foremost constraints associated with the use of Decision Support Systems. Furthermore, the performance metrics adopted revealed that all the selected models performed satisfactorily. However, the performance of the Decision Support Systems is better than that of the Hedonic Pricing Methodology. The study recommended the use of Decision Support Systems for better valuation accuracy.