Strategies for construction hazard recognition
Case 2 description Construction of a power plant project
Download 2.75 Mb. View original pdf
|
- Navigate this page:
- Case 2 results and analysis
| Case 2 description Construction of a power plant project The second project involved the construction of a natural gas power generating plant located in the southern United States. The project scope was $550 million and, during the site visit, the project was 8% complete and had accumulated about 72,850 worker-hours. The site employed 9 active work-crews during the visit, consisting of 78 craftsmen and supervisors. Similar to Study 1, the management devoted three safety personnel to conduct the study with a total of 56 active years of involvement in the occupational safety profession along with one site-based researcher. Subsequently, three crews (civil, plumbing and piping, and equipment operators) were randomly selected for shadowing. The protocol implemented in study 1 was closely followed both during the testing and analysis phase to ensure consistency across sites. Case 2 results and analysis The results of Study 2 are presented Figure 4 and Table 4. As shown, Model I was appropriate for crew 1 and 3 based on the model comparison statistic indicating both a level change and slope change in hazard recognition and communication performance. For crew 2, Model II 102 indicating only level change in hazard recognition and communication performance was more appropriate. That is, each crew exhibited a level change improvement, while only crew 1 and 3 revealed a slope change trend in hazard recognition and communication. As shown, all assumptions of homoscedasticity and normality of error variance were satisfied, suggesting the validity of inferences. Crew 1 revealed a level change improvement of 15% (p=0.002) during the seventh time period when the intervention was introduced with a slope change coefficient of 0.910, implying an improvement of about 1.26% (0.352+0.910) with every subsequent time period. Similarly, crew 3 showed a level change improvement of 18% at the eleventh time period when the intervention was introduced to the crew and an improvement of about 1.65% (0.357+1.293) for every subsequent time period. On the other hand, crew 2 only reveled a level change improvement of 23% (p<0.001). The weighted overall level change computed using equation 3 suggests an improvement of 20% and a slope change of 1.139 for each sequential time period. Similar to Study 1 each crew in Study 2 exhibited a statistically significant improvement in hazard recognition levels in the corroborative two-sample t-test for independent measures using photographs of construction scenarios. Crew 1 revealed an increase of 29% (p<0.001), while crew 2 and 3 revealed 35% (p<0.001) and 25% (p=0.001), respectively. |