Use this URL to cite or link to this record in EThOS:
Title: Construction demand modelling : a systematic approach to using economic indicators and a comparative study of alternative forecasting approaches
Author: Goh, Bee Hua
ISNI:       0000 0001 3501 6115
Awarding Body: University of London
Current Institution: University College London (University of London)
Date of Award: 1997
Availability of Full Text:
Access from EThOS:
Access from Institution:
The published literature abounds with evidence of a close relationship between the construction industry and the national economy. This study reinforces the strength of this relationship by proposing the use of economic indicators to model demand for construction. Alternative forecasting approaches are applied, comprising both traditional and state-of-the-art techniques. The aim is to establish the most theoretically significant and statistically adequate indicators, and the most accurate forecasting technique for modelling and predicting construction demand. A systematic approach is proposed to identify and select economic indicators that relate to demand for construction. It involves four distinct stages and they are: (1) theoretical identification: (2) data collection and pre-processing; (3) statistical selection; and (4) usage. This stage-by-stage process is illustrated on residential, industrial and commercial-type construction in Singapore. The findings confirm that demand in the construction industry is significantly related to a wide range of economic measures. A comparative study of regression and non-regression approaches of forecasting is earned out using Singapore's residential sector as a case-study. The techniques include the Multiple Linear Regression, the Multiple Log-linear Regression, the Autoregressive Non-linear Regression Algorithm and the Artificial Neural Network (ANN). Seven economic indicators have been selected to build the demand models, and they are: Building tender price index; Bank lending for housing; Population size; Housing stock (additions); National savings; Gross fixed capital formation for residential buildings; and Unemployment rate. Quarterly time-series data over the period 1975 - 1994 are used. Several conclusions are drawn. Firstly, non-linear methods produce more accurate forecasts. Secondly, the Multiple Log-linear is the most accurate regression technique. Thirdly, the ANN technique, a non-regression approach, performs outstandingly better than the regression approach. Keywords: Demand, economic indicators, forecasting, regression, artificial neural network.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available