*2.1. Study Design and Randomization*

We conducted a cluster randomized control trial on ten (five matched pairs) commercial construction sites (five intervention; five control) across the Boston metropolitan area, Massachusetts, United States between 2014 and 2015. Construction sites were matched within each general contracting company that agreed to participate in the study. Each pair of worksites was matched based on approximate size, scope, and phase of construction. This ensured that each matched pair consisted of similar organizational and worksite factors, such as similar existing company health and safety management systems [36]. Within each matched pair, one site was randomly assigned to either an intervention or a control group. A blocked randomization sequence was generated using a web-based random number generator by a member of the research team, who then allocated the pairs to either intervention or control. All workers within a specific worksite received the same intervention (or control) as allocated. Randomization and allocation of randomization sequence occurred as soon as two construction sites within a general contractor agreed to participate regardless of their assignment to control or intervention. The intervention groups received the intervention, ARM, whilst the control group received no intervention. Due to the pragmatic nature of the intervention, neither interventionists nor participants were blinded.

#### *2.2. Recruitment and Eligibility*

Construction sites were recruited through construction site owners and general contractors. To be eligible to participate in the study, worksites had to be in operation for 4 months or longer, and have 30 or more workers. Before the intervention commenced, a recruitment meeting was conducted with each site owner or general contractor to provide an overview of the study, programmatic activities, and to obtain leadership commitment. These recruitment meetings were conducted by the intervention primary investigator (J.T.D). Once a general contractor agreed, additional meetings with the leadership of each of the selected sites provided further leadership commitment and agreement for the study to take place on their sites.

Workers were introduced to the study by the research team and the general contractor safety manager, at a "safety stand-down" or toolbox talk. During the study, all new workers on a site were oriented to the study at their new-worker onsite safety orientation. Individual construction workers were surveyed within each site after a study launch meeting and at new-hire safety orientations for those workers who started after study commencement. Surveyed workers at each site were eligible if they were aged 18–65, and were English literate. All surveyed workers self-nominated and provided verbal consent during the survey process

All subjects gave their informed consent for inclusion before they participated in any data collection activities. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Harvard T.H. Chan School of Public Health Institutional Review Board (IRB-13-1948).
