Du, J., & Gunhan, S. (2014, June), Developing a Statistical Model for Building Settlement Prediction  Paper presented at 2014 ASEE Annual Conference & Exposition, Indianapolis, Indiana. 10.18260/1-2--20280

\bibitem{asee_peer_20280}
  Du, J., \& Gunhan, S. (2014, June), \emph{Developing a Statistical Model for Building Settlement Prediction}
  Paper presented at 2014 ASEE Annual Conference \& Exposition, Indianapolis, Indiana.
  10.18260/1-2--20280

Jing Du and Suat Gunhan.  "Developing a Statistical Model for Building Settlement Prediction".  2014 ASEE Annual Conference & Exposition, Indianapolis, Indiana, 2014, June.  ASEE Conferences, 2014.  https://peer.asee.org/20280 Internet.  25 Apr, 2024

\bibitem{asee_peer_20280}
  Jing Du and Suat Gunhan.
  "Developing a Statistical Model for Building Settlement Prediction".
  \emph{2014 ASEE Annual Conference \& Exposition, Indianapolis, Indiana, 2014, June}.
  ASEE Conferences, 2014.
  https://peer.asee.org/20280 Internet.  25 Apr, 2024

@INPROCEEDINGS{asee_peer_20280
author = "Jing Du and Suat Gunhan"
title = "Developing a Statistical Model for Building Settlement Prediction"
booktitle = "2014 ASEE Annual Conference \& Exposition"
year = "2014"
month = "June"
address = "Indianapolis, Indiana"
publisher = "ASEE Conferences"
note = {https://peer.asee.org/20280}
number = {10.18260/1-2--20280}
}

TY  - CPAPER
AB  -              Developing A Statistical Model for Building Subsidence PredictionBuilding subsidence is the downward movement of the ground supporting the buildings. It isvery common in all types of construction especially high-rise buildings and once happened, isvery risky to the occupants. Most building codes have specific requirements on subsidencetolerance. As a result, construction managers, at the early phase of a project, should be able topredict the possible subsidence based on information available including design documents andsoil conditions, to calculate the risk. A variety of computational/empirical models can be appliedto predict building subsidence; however these models are either too complicated to non-engineering practitioners, or inaccurate in many cases.This paper aims to develop a statistical model to predict the subsidence of high-rise buildings. Inthe selection of a proper statistical method two criteria have been highlighted: (1) the methodshould be able to yield results that are accurate enough for practical applications; and (2) themethod should be simple enough for industry practitioners and construction students. This paperproposes the use of regression analysis, which is often utilized to explore the relationshipbetween a set of explanatory variables (X’s) and the observation, or response variable (Y). Inorder to build the model, a dataset was established which contains 33 actual high-rise buildingconstruction projects in China, built from year 2002 to 2008. The response variable (Y) is“Building final subsidence (cm)”. Thirteen explanatory variables (X’s) were initially selectedupon a survey to the project managers and relevant building codes including such as “Number oftower stories”, “ Height of tower”, and “ Type of ground soil” etc. The t-tests identified eight ofthem to be significant to the model and therefore were finally selected. The developed predictionmodel was tested against the empirical formulas and its better predictivity has been confirmedwith a R2 of 0.951.The modeling process has also been tailored into a lecture to teach construction students how toapply advanced statistical methods in construction management and engineering area. The stepscovered include preliminary analysis, model selection procedure, model transformation such asBox-Cox method and stepwise method, and model diagnosis and evaluation. Students were alsotaught how to use Excel to perform regression analysis for construction management andengineering problems, and how to interpret the results. It was found that the regression analysisis easy to understand and always has satisfactory performance in the construction managementand engineering area.
AU  - Jing Du
AU  - Suat Gunhan
CY  - Indianapolis, Indiana
DA  - 2014/06/15
PB  - ASEE Conferences
TI  - Developing a Statistical Model for Building Settlement Prediction
UR  - https://peer.asee.org/20280
DO  - 10.18260/1-2--20280
ER  -