Particulate matter (PM) is an important exposure risk in society (WHO, 2013), as well as in various workplaces, for example in roadwork companies (Meier, Cascio, Danuser & Riediker, 2013; Van Deurssen, 2015). People in these companies regularly inhale the small particles of PM, especially ones with a diameter smaller than 2.5 µm (PM2.5), resulting in potential detrimental health effects (Hänninen and Knol, 2011; Strak, 2012). These effects may include cardiovascular and respiratory diseases, such as lung cancer and bronchitis (Hänninen and Knol, 2011). The effects of PM exposure are estimated to cause around 800,000 annual deaths worldwide (Anderson et al., 2012). Protection against PM involves such measures as sprinkling water, respirators, filtering systems and ventilation (Heederik et al., 2009; Uchiyama, 2013). Research shows that personal protection against PM, mainly in the form of various types of respirators, has a profound effect on PM reduction; however, not all workers that are exposed to PM use them appropriately (Liu, Noth, Eisen, Cullen & Hammond, 2018).
The protection motivation theory, or PMT (Rogers, 1983), distinguishes two processes that influence the motivation to protect against risk. These processes are threat appraisal, which is the perceived expected risk subtracted by the benefits of risky behavior, and coping appraisal, which is the perceived efficacy of protective behavior subtracted by its cost. In general, a higher threat appraisal and a higher coping appraisal lead to a higher tendency to protect oneself against a certain risk. However, in some cases, higher threat appraisal might be counterproductive, and cause people to ignore the message (Goldenbeld, Twisk & Houwing, 2008; Ruiter, Kessels, Peters & Kok, 2014). This may be explained by fear eliciting a maladaptive response as people avoid the risk communication message rather than the risk itself (Rogers, 1983). However, not all researchers agree that these counterproductive effects exist, and some say there is simply a cap on the benefits of threat appraisal (Tannenbaum et al., 2015). Either way, these factors should be taken into account when designing a risk communication material.
In earlier research (Stege et al., 2019), we assessed specific information needs of employees in roadwork companies concerning PM exposure. We did this by means of a mental models approach. Mental models can be defined as “personal, internal representations of external reality that people use to interact with the world around them” (Jones et al., 2011). In risk communication and risk perception research, the mental models approach is a systematic way to map these representations of a risk (that is about sources, properties, exposure, effects and mitigation options), and to contrast the representations of various groups of people, such as experts and non-experts (Breakwell, 2001; Morgan, Fischhoff, Bostrom & Atman, 2002). The concepts of threat appraisal and coping appraisal from the PMT model mentioned earlier resemble the various aspects of mental models of risk. That is, beliefs about sources, hazardous properties, exposure and effects of a certain risk are closely linked to threat appraisal, and beliefs about mitigation methods can be linked to coping appraisal.
After mapping the mental models of various groups, the differences between them are used to identify information needs in risk communication. This way, risk communication can alleviate common misconceptions and answer common questions about the subject matter (Slovic & Weber, 2002). The mental models approach has been used in a wide array of risk-related subjects resulting in suitable risk communication tools, ranging from flood prevention to cigarette smoking (Riley, 2014).
The mental models approach in our previous study (Stege et al., 2019) yielded a scientific and an employee mental model for particulate matter. The scientific mental model was extracted from literature on PM, and corroborated by interacting with experts in the field. The employee mental model was erected after conducting 22 semi-structured interviews with employees in the roadwork sector. An overview of the main differences between both of these mental models can be found in Table 1.
Subject | Scientific mental model | Employee mental model |
Properties | PM is usually invisible | It is unclear whether PM is visible or not |
It is not possible to smell PM | It may be possible to smell PM |
Although small particles of asbestos could technically be considered PM, asbestos is generally thought of as something different | Asbestos may be one of the compounds in PM |
Black carbon, metals, silicium and rubber are important constituents of PM | - |
Particle size is most often defined in terms of PM10, PM2.5 and PM0.1 | - |
PM consists of solid particles | - |
Sources | - | Sand and dirt roads cause PM |
There are natural sources of PM, such as sea salt, which don’t cause adverse health effects. | - |
Effects | - | PM exposure may cause headaches and nausea |
PM exposure is associated with cardiovascular disease, even more so than with respiratory disease | (Almost) no mention of cardiovascular disease; only attention for respiratory diseases |
PM causes about 800,000 annual premature deaths worldwide. | - |
PM is also an environmental risk (for example due to acid rain or nutrient depletion). | - |
Prevention | There is an occupational hygiene strategy that involves a four-level hierarchical model, which should be followed to reduce PM exposure. | There are a large number of prevention methods (sprinkling water, respirators, …) that could be used to reduce PM exposure. |
Education and empowerment | A viable education system improves safety culture and willingness to protect against (exposure) risks. | The current education system could be improved; it is often too ritualistic and repetitive, and not everyone is involved with the process. |
One question that comes to mind is how to convey quantitative risk information about health effects and exposure. Research recommends using a so-called ‘X in 100’ format to convey the potential health effects in a population (Trevena et al., 2013; Visschers, Meertens, Passchier & De Vries, 2009), as percentages alone may confuse the reader and lead to false interpretations. These ‘X in 100’ formats are generally preferred by respondents to similar formats such as ‘1 in X’ (Visschers et al., 2009). In general, visually enhancing risk information with graphs tends to be more effective than simply providing verbal or numeral information (Fischhoff, Brewer & Downs, 2011; Lipkus, 2007). Our own experience in an earlier study was that blue-collar workers tend to find graphs about workplace exposure interesting and insightful (Bolte et al., 2018).
Nevertheless, graphs can also be inadvertently misleading; an example of this involves participants judging a cardiovascular risk from a bar chart as relatively low compared to 100%, even though experts would say that the risk is quite high (Damman et al., 2015). Therefore, it is imperative to choose an appropriate format. Specifically, when considering the number of individuals affected in a population, a ten-by-ten matrix of human icons may be used to convey a percentage (Lipkus, 2007). Visualizations such as these help reduce several biases, including framing effects and denominator neglect (Trevena et al., 2013), although their effectiveness is not explained by an improvement of exact knowledge about the risk; only ‘gist knowledge’ appears to be increased (Etnel et al., 2020).
In our situation, we would like to give a rough but accurate estimate of the health risk of PM, in order to induce an accurate representation of the risk. Exact numbers for the amount of work-related deaths due to PM are unknown, as the earlier mentioned 800,000 deaths per year worldwide applies to all people in general, without any indication how many of these deaths are work-related (Anderson et al., 2012). However, an estimate can be generated from a factsheet about hazardous substances at work (Arboportaal, 2018), which mentions that one million people in the Netherlands are exposed to one or more hazardous substances at work, and 3,000 of those people die each year. Although these may be various types of substances and not just PM, the most important substances mentioned are all a form of PM, such as diesel emission (Heederik et al., 2009) or quartz (Van Deurssen, 2015).
In this study, we considered the aforementioned recommendations about contents as a basis for developing and testing new educational material about workplace PM exposure. Next, we will present the method by which we developed a new educational material for blue-collar workers about PM. Furthermore, we consulted experts on risk communication, particulate matter, or both, as well as blue-collar workers, in order to inquire about the usability of our educational material. The question we would like to answer is, ‘How do stakeholders perceive the usability of a mental model-based educational material about workplace PM exposure?’