Abstract:
Construction industry is complex in its nature since it involves large number of
stakeholders. In Ethiopia the number of road projects is increasing from time to time due
to the emphasis given to the sector. However, it has become difficult to complete projects
in the allocated cost, time and quality. The main objective of this research was to develop
forecasting model that helps construction stakeholders to precisely predict construction
performance of a road construction projects by using multi-variate regression technique.
Detailed literature review was done and 61 factors that had tangible impact on
construction performance were listed and grouped in 7 groups namely: Cost, Time,
Quality, Productivity, Client Satisfaction, Health and safety and Innovation and Learning.
A questionnaire survey was conducted on ongoing Federal road construction projects in
the Central Region. Total population sampling technique was used to collect the data since
the population to be studied is limited. Based on the result of the survey, from the 61 factors
9 were selected as most significant factors based on their RII value. These nine factors were
then subjected to Principal component analysis method in order to reduce their number,
and four factors (Delay in payments, Cash flow Problem, Resource Availability and
Sequencing Practice according to schedule) were factored out for the model formulation
process. From the collected data 80% were used in the training set or model formulation.
Multi-Variate regression was done using SPSS to generate the forecasting model. An Enter
selection method was used in developing the model, which resulted in R
2
value of 84.9%
which indicates a good predicting capability of the model. Finally, the developed model
was validated by using the remaining 20% of collected data and resulted in small amount
of mean average percentage error (MAPE=2.98%). The developed predicting model helps
construction stakeholders mainly contractors involving in road constructions in the initial
and construction stages by providing them a predicting formula. And also Contractors are
advised to give serious attention for the factors listed out in this paper as they have a great
impact in affecting construction performance.
Keywords: Construction performance, Model formulation, Multi-variate regression