Exponential, gamma, and power law distributions in information flow on a construction site

This paper examines the statistical distributions of interarrival and response times for construction-site correspondence. Data from a number of construction projects are analyzed and hypotheses are proposed and tested to link the probability distribution to the type of correspondence. Although the commonly assumed exponential distribution of interarrival times is found to be accurate for all types of correspondence, the response times for different types of correspondence follow different distributions depending on the type of correspondence. In particular, a power law relationship is observed between incidence frequency and response time for requests for information. Knowing the statistical distributions for a class of events helps managers forecast future work and manage risk. Simulation models used by practitioners and researchers for various project management goals are also improved by incorporating appropriate statistical distributions for generating various events.