It is common for occupational health and safety (H&S) research to discuss the dynamic environments of high-risk workplaces, including mining, construction, forestry, oil and gas, and transportation. Although the ongoing presence of hazards is real, in some ways, the risks and hazards present in a workplace are fairly static. For example, hazards consistently identified at mine operations include slips, trips, and falls, as well as electrical issues, guarding, or noise and dust exposure [1–3].

Despite a specific hazard or group of hazards being constant in many mine environments, workers often are not. Rather, general occupational research has found that the workforce is constantly changing due to issues such as employee turnover, new people coming in, or job reassignment [4]. These workforce issues are particularly prevalent in industries that have suffered a higher turnover for the past number of years, such as mining, which will continue as baby boomers retire [5]. In response, understanding the knowledge, perceptions, and behaviors of workers, based on any individual differences, may be important for identifying the likelihood of preventing and introducing risks on the job for certain subgroups of workers. In this case, we focus on workers’ experience levels in different job tasks, with their company, and within the industry as individual factors of interest.

To date, even the largest meta-analyses have revealed mixed findings pertaining to the relationship of job tenure and job performance, with more attention being focused on the risks that new employees bring to the job site [6]. Specifically, nearly a century’s worth of research has consistently shown that workers new to their profession are more likely to be injured on the job than those with longer job tenures [7–9]. A commonly referenced landmark study by Leigh and colleagues [8] asserted that length of time on the job (tenure) was negatively associated with accidents (i.e., new employees had a higher accident rate). Similarly, research has also confirmed that job tenure can be positively related to task performance over time [10].

Older studies have shown similar results in coal mining, specifically, where job tenure was found to be a significant factor in accidents that occurred between 1976 and 1977 [11]. These researchers found that 40% of reported injuries had occurred during workers’ first year of employment. More recent research by Groves and colleagues [12] used data from the Mine Safety and Health Administration (MSHA) between 1995 and 2004 to conclude that 28% of injuries and 31% of fatalities occurred among employees who were in their first year on the job. MSHA is the regulatory entity for U.S. mining, mandating specific Codes of Federal Regulations (CFRs) which organizations must follow. Similarly, Kecojevic and colleagues [13] analyzed the relationship between work-related fatalities and worker experience to find that 44% of workers killed in equipment-related incidents had less than five years of mining experience.

Although these numbers seem convincing, there are other factors to consider. First, the statistics can be recontextualized. For example, Kecojevic’s results [13] could communicate that 56% of those workers killed in equipment-related incidents had more than five years of mining experience, indicating that this more experienced population is overlooked when considering specific risk management interventions. Specifically, many studies do not examine the spectrum of the workforce and often focus on workers who tend to be inexperienced (e.g., under one year on the job). One research study showed a strong inverse relationship between job tenure and injury claim rates, but job tenure only ranged from 0 to 13 months [14]. Importantly, these inferences are all made based on reported incidents, whereas we know that underreporting is a problem in high-risk occupations [15]. In light of these issues, a more diverse research portfolio that includes all levels of experience, from novice to veteran, is needed to better understand this phenomenon.

As this review demonstrates, there is still much disagreement on the impact of workers’ demographic factors, specifically experience, on performance outcomes. Therefore, more recent and even self-reported data that offer the ability to study a variety of experience factors, as well as to determine accurate occurrences of incidents, are necessary to uncover any trends in potential indicators of incidents, if they exist. To that end, the National Institute for Occupational Safety and Health (NIOSH) researchers analyzed several quantitative datasets that were collected by working with miners who worked in either underground coal, surface stone, sand, and gravel, or metal/nonmetal to reveal any underlying trends among differing levels of experience. These case studies include several sectors of the mining industry. The mining industry remains an important part of economies both domestically and internationally. Regardless of the sector (e.g., coal, metals, minerals, etc.) and whether materials are extracted on the surface or underground, the working conditions are consistently acknowledged as physically demanding, and workers are exposed to numerous risks [16]. In 2007, National Public Radio [17] interviewed mineworkers to better understand their job and work environment, and a common theme among employees was that the work is “dark, dirty, and dangerous.” Due to the numerous hazards present both underground and on the surface, research to understand ways to improve the H&S workers through protective technologies and evolving risk management processes remains important.

Mineworkers completed respective surveys over a two-year period between 2016 and 2018. These data, collected in three separate case studies, used time in job, time at the mine, and time in the mining industry as predictor variables in an attempt to find associations with workers’ perceived knowledge, confidence, and job compliance. These specific case studies provide results that show the necessity of understanding how experience, on multiple levels, can impact worker perceptions and performance in order to improve hazard recognition, compliance to safety rules, and if necessary, rapid response in the event of an emergency. Although the results in this study are specific to mining, the trends are still relevant to any high-risk industry to help develop and implement aspects of a tailored risk management program to segments of the workforce at opportune time periods. Whether focusing on the development and implementation of more accurate hazard recognition and training, adequate preparation and effective response during a mine emergency, or improved safety climate through targeted leadership and engagement in H&S, all are efforts that can benefit from more information about the workforce at varying points in their careers.

Prior to the start of each respective data collection, NIOSH researchers obtained Institutional Review Board (IRB) and Office of Management and Budget (OMB) approval. In total, 3428 surveys were completed with mineworkers over a two-year period. These data, collected in three separate case studies, use time in job, time at the mine, and industry tenure as predictor variables in an attempt to find associations between workers’ varying experience and performance outcomes of interest (i.e., hazard recognition, self-escape confidence, and compliant behavior). Each of these case studies are presented in turn and utilize similar analysis methods to convey the results related to experience. Regarding how experience was measured within each case study, the categories measured were under one year, 1–5 years, 6–10 years, 11–15 years, 16–20 years, and more than 20 years. These categories are consistent with the demographic breakdown included in NIOSH’s Quality of Worklife Questionnaire and General Social Survey. These surveys have generated reports since 1972, exhibiting reliability and consistency in demographic breakdowns. Additionally, MSHA asserts that a novice worker has under one year of experience, further supporting our justification for the inexperienced group.

Prior to running analyses within each case study, an assessment of the normality of the data was completed using either a visual inspection (i.e., Case Study 1) or the Kolmogorov-Smirnov test (i.e., Case Studies 2 and 3). For Case Studies 2 and 3, the Kolmogorov-Smirnov test rendered a result of less than 0.05, showing a deviation from normal distribution. This is normal and even common for large datasets [18]. In response, a non-parametric option, the Kruskall-Wallis, was used as an alternative to a one-way between groups ANOVA for all three case studies. Similarly, in order to understand exactly where each significant relationship existed, Mann-Whitney tests were completed for each group to further investigate differences between the various experience groups and outcomes of interest. To better account for a potentially inflated type I error, due to multiple testing, the researchers adopted a more stringent significance value of 0.01 as an alternative to Bonferonni tests for each potential relationship. Further, we report the skewness and Shapiro-Wilks test of normality to demonstrate the non-normal distribution of the results, further justifying the use of nonparametric tests.

Now that all three of the case studies and their respective results have been explained, it is possible to compare the results across cases to determine trends among various types of experience. The case studies, each presented separately, have slightly different variables, but render some similar and deviating trends that are worth noting. Specifically, these trends can not only be used to inform and improve aspects of an occupational risk management system, but also to tailor practices, activities, and resources for at-risk worker segments in the industry to reduce incidents for workers of varying experience levels. If the results reveal anything, it is first, that there are several types of experience that researchers, practitioners, and management must account for when determining how to improve the execution of specific risk management processes in the workplace. More specific implications are discussed below.

It is an obvious limitation that we are drawing upon patterns using three varying datasets that are all based on behavioral lab or self-reported data. If researchers were able to coordinate data collections and specific instruments to a more rigorous degree, stronger relationships and results may have emerged. For example, being able to track the same groups of people to compare how they perceive and process risks that stem from routine and non-routine work practices, would be an area of future exploration that could occur with a common sample. However, the results for each separate case study all support findings in previous studies, supporting the validity of all three studies and their potential influence on future organizational research. In our case, synthesizing the results of the three studies together allowed for some sort of triangulation, particularly among the second and third case studies, which provide support for increased communication and on-the-job experience and self-assessments that can support the ongoing development and maintenance of procedural knowledge. However, future research should aim to better coordinate big data collection efforts with overlapping populations in order to draw more cause-effect results. In addition, such data should be linked to lagging indicator data [67] to better support the conclusions made throughout this paper. The current case studies did not collect accident or fatality data so these concrete links cannot be made.

Despite all the limitations, this paper was able to contribute to the occupational H&S literature in several ways. First, we were able to show that organizations need to consider worker factors other than age and blanket experience as possible predictors of worker performance. The main takeaway is that even if longer job tenure is associated with safer job performance, as shown in Case Study 1, this may not be the case as workers’ tenure increases at the mine site and within the industry as an overall career path. More specifically, our results demonstrated that communication with more intermediate to veteran workers (i.e., 6–20 years) is necessary to maintain worker engagement and execution of the organization’s health, safety, and risk practices.

In summary, these three case studies show the value of being able to identify various levels of workers’ experience, and that it is important for an organization to continually reengage with workers about specific safety issues relevant to their job tasks, at their mine, and within their mining sector. In doing so, managers may be able to parse out important differences in their workforce’s knowledge, attitudes, and behaviors based on varying levels of experience, in order to reduce negative consequences for managers and their organization while improving the knowledge, motivation, and overall resources for the workforce.