Psychosocial hazards or risk factors can be defined as "those aspects of work design, and the organization and management of work, and their social and organizational contexts, which have the potential for causing psychological or physical harm" (Cox and Griffiths, 1996, 129-130). This research proposal provides an analysis of problems with current measures of psychosocial hazards, and then investigates the theories that underlie how work events lead to emotional and physical reactions. This research proposal then proposes to help further the theoretical understanding of the interaction between stress and health reactions. Specifically, this proposal attempts to continue to determine the nature of stresses that lead to harms, within the context of Siegrist and Peter's earned reward imbalance model. This study will attempt to determine the core expectations (as defined within Siegrist and Peter's earned reward imbalance model), and hypothesizes these expectations include immediate expectations of specific salary requirements, wage increases, working conditions, and social and emotional feedback, as well as longer-term expectations about status.

A brief discussion of stress and its relationship to harm may be helpful in setting the context for this proposal. Increased stress can be caused by a number of factors. These can include insufficient personal time, high time pressures including deadlines and crisis at work, major consequences for work actions or errors, and a high rate of work change, including organizational and technological changes (Macdonald, 2003). Further, Britain's Health and Safety Executive (HSE) notes that stress hazards can include "lack of control over the way you do your work, work overload (or underload), lack of support from your managers, conflicting or ambiguous roles, poor relationships with colleagues (including bullying), or poor management of organizational change."

There is an established link between psychosocial hazards, stress, and physical injuries such as musculoskeletal injuries. Specifically, monotonous work coupled with time pressures and a rate of high perceived injuries are associated with musculoskeletal symptoms. Such symptoms are also closely linked to a lack of social support by colleagues, and low job control, with stress acting as an intermediary between psychosocial hazards and musculoskeletal symptoms (Bongers et al., 1993). Further, high job strain (as determined by the Karasek and Thorell demand-control model) has been linked to back injuries (Myers et al., 1999). A review by Devereux and Buckle (2000) confirmed this existing link between physical symptoms and stress, and noted that neck-shoulder pain and lower back pain were predicted by work-related stress.

One way that is commonly used to attempt to reduce workplace harm due to psychosocial hazards is through attempts to measure these hazards. It is thought that once these hazards are identified, organizations can word to reduce such hazards, and thus reduce eventual harm to employees.

However, currently research on questionnaires designed to measure psychosocial hazards is neither reliable nor valid. Existing approaches to the measure of psychosocial hazards are largely based on stress questionnaires, despite their numerous potential problems. These questionnaires are presumably designed to quantify factors that cause stress in the workforce (Rick et al., 2001).

Rick et al. (2001), in a report for The Health and Safety Executive (HSE), note that research on commonly used hazard-measuring questionnaires does not produce consistent measures, and thus is generally not reliable. Further, research on these questionnaires does not necessarily measure what it is designed to assess, indicating problems with validity (Rick et al., 2001).

Among other problems, the quality and quantity of evidence reported in the research of different questionnaires was limited. Studies commonly suffered for inconsistent reporting of data, and often lacked internal analysis (Rick et al., 2001).

Overall, these limitations make it impossible to recommend specific questionnaires as useful measures of psychosocial hazards (Rick et al., 2001). While the stress questionnaires that were studied were demonstrably good at identifying specific hazards, there was little evidence as to whether the questionnaires measured the hazards that played a role in psychosocial risks (Rick et al., 2001). Writes the Australian Chamber of Commerce and Industry, "If hazard identification, and ultimately risk assessment are to be successful, then it is essential that the measures used for these are reliable and valid. What this report shows us, overwhelmingly, is that they are not."

The lack of good information about the reliability and validity of current hazard-measuring questionnaires demands that businesses rethink how they measure workplace stress. Rick et al. (2001) note that organizations should consider devising measures of stress in the workplace that go far beyond self-reporting questionnaires....

These measures of stress should be specific to certain jobs and roles in the organization, based on best practice scenarios, focused on local knowledge and context, and be incorporated into a risk management framework (Rick et al., 2001).
Further, Rick et al. (2001) note that organizations should rely upon more than a single, specific measure of work-related stress as a reliable measurement of risk assessment. Information from a variety of sources may be useful, and include job descriptions, observations, task analysis, and reports of harmful incidents and the reports' conclusions about the cause of such incidents.

Notes the Australian Chamber of Commerce and Industry, "If employers cannot assess harm with certainty, how can the validity of hazard prevention be measured? Risk assessment can only be successful if the measures used for hazard identification are reliable and valid. Workable and effective specification and quantification of harm has so far eluded even expert researchers, so employers face an impossible task, uncertain which of the many varied stress theories and interventions has the answer."

One of the greatest contributors to current problems with the validity of psychosocial hazard measurements can be traced back to limited understanding of the theory that underlies how work events lead to emotional and physical reactions. If we do not understand specifically what we are measuring, and why and how we are measuring it, then it is perhaps not surprising that current psychosocial hazard measures are problematic. An investigation into theories within the stress field, and how work events can lead to emotional and physical consequences can be helpful in helping to determine a psychosocial hazard measurement tool that is based on sound theory. Specifically, Rick et al. (2001) suggest that other methods to assess hazards could include "observations, task analysis, job descriptions, and reports of harms and what these may tell us about hazards" (p. 83).

Essentially, many of the problems with the measurements noted above can be linked to problems with our understanding of the theory that underlies how work events lead to emotional and physical reactions. If our hazard measurements are not based on sound understanding of the theory that underlie such events, the issues with the validity and reliability of psychosocial hazard measurements, as discussed previously, will continue to be a problem. It is crucial that psychosocial hazard measures must be both valid and reliable, and therefore based on theoretically sound understanding of stress. As such, this research proposal attempts to create a viable theoretical background to underpin an approach to psychosocial hazard measurement that is based upon a sound theory and understanding of the interaction between stress and health reactions.

Note Rick et al. (2001), "there are relatively few theories about the nature of stress which can be used to guide the measurement of hazards" (p. 31). In contrast, there seems to be adequate evidence that describes the relationships between hazards and harm but little theory that adequately explains such relationships.

One exception is Siegrist and Peter's (1996) reward imbalance model. As Siegrist (1996) notes, this model claims that low workplace rewards (in terms of career opportunities, money, and esteem), in association with recurring high effort, creates strains that are associated with adverse physical and mental outcomes. A recent study of 11,636 employed Dutch men and women (de Jonge et al. 2000) indicates that employee well-being is reduced under conditions of reward imbalance.

Another theory of the nature of stress is Karasek's Job Demand-Job Control model (1979), which postulates a relationship between job demands, job controls, wages and stress. This theory "states that the greatest risk to physical and mental health from stress occurs to workers facing high psychological workload demands or pressures combined with low control or decision latitude in meeting those demands" (Schall, 1998). A recent study by Shen and Gallivan (2004) notes that "negative consequences are moderated by the amount of autonomy that employees experience in their work," while a 1994 study by Schall and a 2002 study by Kivim ki et al. link job strain to cardiovascular disease. While the link between low control, high workload, and harm seems to support Karasek's Job Demand-Job Control model, Rick et al. (2001) note that Karasek's theory does not "suggest ways in which (hazard) measures can or should be developed" (p.34).

One interesting approach to furthering the theoretical understanding of the nature of stress would be to try to combine features of Karasek's Job Demand-Job Control model (DCM) and Siegrist and Peter's effort reward imbalance model (ERI). However, a 2004 study by Calnan et al. suggests that "there is little evidence to support a combined model of work characteristics. The ERI appears to be the stronger of the two models although the DCM has…