Search

Construction labor productivity remains of great importance due its direct effect on project costs. Numerous parameters (factors and practices) that critically affect labor productivity have been identified in past studies through expert knowledge obtained from surveys. The objective of this paper is to explore whether there is a gap in experts’ knowledge in identifying the critical parameters by comparing their perspectives to the results of data-driven analysis of the parameters and labor productivity field data. This paper presents a methodology for identifying critical parameters using both a factor survey and a data-driven approach. The factor survey approach ranks the critical parameters based on the responses of both project management and trade level personnel on a project. The data-driven approach ranks the parameters based on their degree of influence on productivity through filter feature selection on data collected from the actual project. Results of the comparison of factor rankings from the project management perspective, trade perspective, and data-driven approach indicate a major discrepancy between the expert perspectives and the data-driven results, suggesting a need for verification of expert-based results with additional field studies of factors affecting labor productivity.

Work sampling has been used to indirectly measure crew productivity. Although favoured for being less costly, easy to adopt, and able to provide quick information, previous work sampling studies have not gone beyond identifying the direct, support, and delay proportions of activities to provide a reasonable estimation of productivity. Research into identifying a relationship between work sampling results and productivity has been limited, and an approach to identify the most productive proportion of direct and support activities for different trades has not been developed. This paper proposes a research framework for crew-based work sampling, supplemented by foreman delay surveys and craftsman questionnaires, to establish a relationship between work sampling and productivity, and to identify the effective proportions of direct and support activities for different trades. The paper describes the development of this framework and illustrates the analysis involved by using case study data. Ultimately, this framework will be used to develop a crew-level productivity analysis model, based on subjective and objective factor modeling, supplemented by work study methods, including work sampling.