Strategies for construction hazard recognition
PRACTICAL CHALLENGES WITH IMPLEMENTING MBT FOR CONSTRUCTION
Download 2.75 Mb. View original pdf
|
| PRACTICAL CHALLENGES WITH IMPLEMENTING MBT FOR CONSTRUCTION Conducting field experiments can be challenging in dynamic and rapidly changing construction environments. The difficulty is further amplified when human-centric studies involving the study of human performance or behavior is conducted. Therefore, not surprisingly, the number of studies that adopt experimental methods in construction research is limited. This is because the researcher is conducted in the natural environment or setting and the researcher has limited control over projects activities, schedules, weather, and the multitudes of confounding variables that can affect the validity of the study. For example, it is unrealistic to conduct true randomized experiments inmost cases. This is because workers from different trades or crews if manipulated randomly to form the treatment group and control group, then the results cannot be generalized for realistic scenarios. More importantly, manipulating the structure of crews is highly disrupting and costly, and therefore not practical. Also, as discussed in the previous section, it is unethical to conducted randomized experiments which deprive certain workers a safety intervention that can reduce injuries. However, as discussed in Chapter 3, 4 and 5, the Multiple baseline testing method allows the researcher to conduct experiments, compute effect size effectively, and establish reliable causal inferences. 167 However, we faced several other practical challenges while implementing the MBT procedure in practice. First, it was necessary to measure hazard recognition performance of multiple crews concurrently. Therefore, it was necessary to recruit additional safety professionals for conducting field observations. Second, it was necessary to calibrate each observer to ensure the reliability and consistency between observations from different individuals. This required that we provide pre-calibration training using example photographs of construction projects, and empirically evaluating reliability measures. In this study, although we did not measure the level of agreement or concordance through formally established methods, we conducted brainstorming sessions after each work period to ensure agreement between the observers. In future studies, using formal methods to measure concordance maybe necessary. Third, the cost and time invested in conducting such longitudinal studies is comparatively high. For example, this study involved travelling to six construction projects located in five different states. Also, it was necessary to spend more than 100 days in the field to gather the necessary data. This challenge was compounded when weather conditions disrupted work activities and the research endeavor. For example, during my visit to New Orleans, the construction site was disrupted by hurricane Issac, and therefore evacuation in the construction site was ordered. Therefore, it was necessary to reinitiate the study after the effects of hurricane Issac subsided and work resumed. Fourth, a significant problem with conducting staggered studies was the possibility of the diffusion threat. In our case, the threat was that the intervention introduced to one crew would diffuse or would be shared with other crews that were still in the control phase. In order to 168 control such treats, the first intervention was provided only after six work periods to the first crew. With this baseline data, it was possible to ensure that the gathered data was not noticeably affected by diffusion. Also we requested of the upper management to provide us access to crews that work at different locations, or have different break-rooms to minimize diffusion threats. Fifth, although we intended to use voice recorders to minimize any observer effects that may result due to research subjects altering their behavior due to the presence of external researchers this was not possible due to disclosure and confidentiality requirements that were imposed by the host organizations. However, other methods were used to the best of our ability as discussed in Chapter 3, 4 and 5 to reduce any observer effects from affecting the reliability of research findings. Sixth, for some strategies, there were issues that hindered optimal use of the devised strategy, For example, as per the protocol of using the HIT board, it was necessary workers to place the board adjacent to where work was being carried. However, there were challenges with transporting the board to different locations within a single day. For SAVES, our first site did not have computers with the necessary hardware capabilities to successfully run the software. Therefore, it was necessary to substitute and purchase additional computers for this purpose. For the SMQM tool, it was impossible to hold safety meetings adjacent to the work locations when noise levels were high or was not adequate or unsuitable to accommodate the whole crew. Other necessary requirements for the successful implementation of such strategies in practice include management support, adequate safety staffing, and worker involvement. |