Improving Occupational Safety and Health Intervention

Thus, unintentional injury represents a serious public health concern, and a theory-driven community, school, and organizational injury prevention technology is needed to improve the health and safety of individuals. THE IMPACT OF OCCUPATIONAL INJURIES Due to the frequency and severity of injuries, the U. S. Department of Health and Human Services has identified injury prevention as a priority for attaining the goals outlined in Healthy People 2000: National Health Promotion and Disease Prevention Objectives (1990). Baker, Conroy, and Johnston (1992) found that injuries occurring on the job due to unsafe (or at-risk) work behaviors remain a significant problem in the U. S. and are a leading cause of unnecessary morbidity.

Every day, an estimated 36,000 employees are injured and 16 are killed (NIOSH, Improving Intervention Effectiveness 2 1998). Moreover, an estimated 7,000 to 11,000 workers die annually with 2. 5 to 11. 3 million employees suffering non-fatal injuries (Leigh, 1995; Miller, 1997). INTANGIBLE EFFECTS Occupational injuries result in 250,000 potential productive years of life lost annually – more than cancer and cardiovascular disease combined (Baker, et al. , 1992; Leigh, 1995). Over and above the traumatic personal consequences experienced by employees and their friends and families due to unexpected industrial injuries and mortalities, there are also critical social and economic consequences to consider.

Although pain and suffering caused by these misfortunes cannot be quantified, the social and economic costs can be estimated. The overall liability of work-related injuries in 1992 has been estimated at $116 billion (National Safety Council, 1993). This figure is an increase from the 1989 estimate of $89 billion and is dramatically larger than the 1985 estimate of $34. 6 billion (Leigh, 1995). These costs include lost wages, medical expenses, insurance claims, production delays, lost time of coworkers, equipment damage, fire losses, and indirect costs (Miller, 1997; National Safety Council, 1998). Even though all of these estimates are enormous, the numbers also indicate that the liabilities of industrial injuries are increasing at an alarming rate.

Today’s estimates show that each year employers pay approximately $200 billion in direct costs associated with injuries both on and off the job. Occupational injuries account for three-quarters of this total or nearly $155 billion annually. This amounts to over $1,400 per work-related injury. The majority of these losses are in the form of insurance premiums for workers and their families and workers’ compensation for days lost from work (Miller, 1997). It is noteworthy, however, that these loss figures likely underestimate the true impact of industrial injuries. Many occupational injuries go unreported and the Occupational Health and Improving Intervention Effectiveness 3

Safety Administration (OSHA) does not use transportation deaths, suicides or homicides (an estimated 50% of all job-related traumatic deaths) to calculate their estimates (Baker et al. , 1992; Leigh, 1995; Miller, 1997; National Committee for Injury Prevention and Control, 1989; Weddle, 1996; Wilson, 1985). The manner in which employees are hurt varies dramatically, and prevention strategies need to address a myriad of environment, behavior and person factors that contribute to each injury (Geller, 1996; Heinrich, 1959; Petersen, 1996; U. S. Bureau of Labor Statistics, 1997). Thus, critically examining and redefining industrial safety research to improve long-term and broad-based impact has important implications for reducing morbidity and mortality and increasing the quality of life among industrial workers.

In order to better understand occupational safety and health and to investigate strategies for improving the effectiveness of industrial interventions for increasing safety-related behaviors, the primary purposes of this research is to compare self-efficacy theory (Bandura, 1997) and the stages of change model (Prochaska, Redding, & Evers, 1997) as a determinant of safety-related behaviors and the impact of a behavior-based safety improvement process. Thus, the goals of this research were to: 1) Develop and evaluate questionnaires to assess employees’ safety self-efficacy and safety stages of change. 2) Administer the newly developed surveys in an industrial setting. 3) Provide the surveyed organization with behavior-based safety (BBS) education/training. 4) Measure employees’ levels of involvement in the BBS process. Improving Intervention Effectiveness 4 5) Compare the predictive effectiveness of self-efficacy vs. stages of change based on the survey results and involvement information.

HISTORY OF OCCUPATIONAL SAFETY AND HEALTH The study of occupational safety and health has been in existence for as long as there have been structured work environments. Hippocrates (460-377 BC), for example, wrote of the harmful effects of an unhealthy workplace on slaves, and Caesar (100–40 BC) was reported to have an officer in charge of the safety of his legions (Pease, 1985; Weaver, 1980). This section traces the history of the various interventions developed to improve workplace safety. The history of occupational safety and health is vast and diverse, and therefore a comprehensive review is beyond the scope of this paper (see Geller, 1996; Guarnieri, 1992; Heinrich, 1959; Heinrich, Petersen, & Roos, 1980; Margolis & Kroes, 1975; Weindling, 1985 for more comprehensive reviews).

However, this section will focus on the major influences (i. e. , government, insurance, engineering and psychology) and pertinent legislation that have shaped occupational safety and health intervention research. THE BEGINNING OF OCCUPATIONAL SAFETY & HEALTH In the midst of the Middle Ages, George Bauer (1492-1555) wrote several books on mining/metallurgy describing several innovative approaches for improving ventilation for workers in mining shafts (Raouf & Dhillon, 1994). And Bernadino Ramazzini (1633-1714), the father of occupational safety and health, also wrote on the safety aspects of mining as well as glass working, painting, grinding, and weaving.

In De Morbis Artificum, or the Disease of Workers, Ramazzini (1713) was the first to document the deleterious effects of working conditions on employees’ health and studied the injury and death rates of many different occupations. Appreciative of the social importance of the progress and economical development Improving Intervention Effectiveness 5 of these occupations, Ramazzini discussed and suggested several preventive strategies for reducing occupational disease and injury (Pease, 1985; Pheasant, 1991; Raouf & Dhillon, 1994; Tayyari & Smith, 1997). Although these early safety engineers did not focus their energies on implementing intervention strategies in the workplace, they certainly laid the foundation for current approaches to reduce occupational illness and injury.

As the machine age dawned with James Watt and Eli Whitney during the late 1700s, employers accepted industrial injuries and deaths as part of the working conditions without considering the economical ramifications. Employees were seen as volunteers, and were plentiful and replaceable (Leigh, 1998). Although the conditions in the early factories were horrendous, with two thirds of the employees being women and children working 12-hour days, people would risk disease, dismemberment and death for employment and a method for providing food for their families. Even if an employee suffered an illness or injury, they would seldom report the sickness because serious or frequent illnesses were cause for dismissal (Heinrich, 1959; Weindling, 1985).

GOVERNMENTAL INFLUENCE As industrial centers grew, the degradation of living conditions increased and the death rate grew. In England, for instance, the first attempt of governmental intervention (1933) began with federally run factory inspections. Table 1 reviews significant milestones toward intervention development in occupational safety and health. The results of the scrutiny by governmental inspectors (most of whom were physicians) had little impact on the health and safety of employees until the mid 1800’s when the Great Factory Act was initiated. _________________________________ Insert Table 1 About Here _________________________________ Improving Intervention Effectiveness 6 The Great Factory Act of 1844 improved England’s factory conditions somewhat, ut employers still saw no economic impact of an unhealthy or a risky workplace. In fact, the families of employees who died on the job had little legal recourse. At most they had their funeral expenses covered by the employer (Heinrich, et al, 1980). In 1880, England passed the Employers’ Liability Act that made it possible for employees, or their families, to sue an employer for damages. This act made the employers more cognizant of the costs of not addressing the safety of their working conditions. However, the family still had the difficult task of proving the employee (or a fellow employee) was not the cause of his own death, was not aware of the hazard, or that the employer was negligent.

Factory inspections and the current laws increased employers’ awareness of occupational safety, but it was not until the worker compensation laws were passed that industry owners finally began to realize the costs associated with occupation injuries. Worker compensation laws covered employee injury regardless of fault; but employees could no longer sue their employers under common law (third party lawsuits were still legal). Seemingly, the worker compensation laws were passed to protect employees. However, they were actually passed to control the number of large lawsuits against employers, and thus enabling a “predictable cost of doing business“ (Leigh, 1998, p. 254). Hence, up to this point the most effective interventions for improving occupational safety and health appeared to be implementation of top-down governmental regulations. As Heinrich, et al. 1980) point out, “Legislation is one process by which government affects safety. Judicial process is another. Together they change the impetus for safety or create a new impetus, and the impetus is defined as time, money and effort” (p. 361). Thus, regulations finally made it cost effective for employers to attend to working conditions that adversely effect employees’ health Improving Intervention Effectiveness 7 and safety, though they were not always in the best interest of the employee (Heinrich, 1959; Heinrich, et al. , 1980; Petersen, 1989; Weindling, 1985; Wilson, 1985). INSURANCE COMPANIES: THE FIRST SAFETY CONSULTANTS Figure 1 depicts the many historical influences of occupational safety and health interventions.

As the worker compensation laws created a need for industries to invest in additional insurance, insurance companies needed to assess their clients’ risks to assign proper rates. Thus, in the early 1900s, insurance companies created inspection departments (see Figure 1 for a historical flow chart of occupational safety and health). The inspectors would visit their policy holders to assess workplace hazards and assign the proper rate (i. e. , underwriting). As these insurance inspectors gained valuable experience in looking for hazards in various industries, these safety consultants became the major impetus in organizational safety and health.

During an inspection, for instance, if the insurance representative found a hazardous situation, he would make suggestions on how the organization could remedy the safety hazard and obtain a lower premium (also to control the insurance companies’ losses). _________________________________ Insert Figure 1 About Here _________________________________ The insurance companies were serving the employer while at the same time trying to control their own losses. Consequently, the only safety concerns addressed by the insurance inspectors were ones currently covered by Worker Compensation laws. Furthermore, once the insurance agent assigns coverage rates, there were several self-serving mechanisms to motivate employers to improve the safety of their workplace. Merit rating schemes (i. e. scheduled rating), for example, rewarded loss control and penalized high worker compensation claims. The Improving Intervention Effectiveness 8 scheduled rating system may have motivated many companies to cover-up or not report certain claims to insurance companies in order to avoid a penalty or keep their current coverage rate (Geller, 1996; Miller, 1997). Whereas it seemed the early insurance companies were striving for a safer workplace, they were instead trying to control their own loss and motivate employers to address only hazards covered by Worker Compensation (Heinrich, et al. , 1980). If fact, most of the insurance inspector’s time and safety materials ent to the larger companies who paid massive premiums, leaving out the mid-sized to smaller organizations. Insurance companies did develop safety guidelines and training materials that made an impact on health and safety. Nevertheless these interventions were guided by current governmental regulations and the need to control loss and not for the safety of employees. TRADITIONAL SAFETY & HEALTH INTERVENTIONS: THREE “E”S OF SAFETY From the early 1900s to the present time, employers and safety practitioners adopted the philosophy of the three E’s (engineering, education, and enforcement) to guide their safetyrelated interventions (Geller, 1996; Guastello 1993; Heinrich, et al. , 1980; Petersen, 1996; Wilde, 1998).

To make a difference in the health and safety of employees, the three Es of safety focus on: 1) developing engineering strategies that decrease the probability of an employee engaging in at-risk behaviors; 2) educating and training employees regarding equipment, environmental hazards, policies and procedures; and 3) enforcing the policies and procedures related to operating equipment, wearing proper personal protective equipment, and handling specific hazardous substances. Improving Intervention Effectiveness 9 ENGINEERING/HUMAN FACTORS The Great Factory Act was passed in Briton in 1844 and by 1850, industrial engineers were improving the physical working conditions by increasing ventilation, improving lighting, and providing guarding for dangerous moving machinery (Grether, 1975; Heinrich, 1959). Industrial safety engineering research has suggested that injuries occur as the result of excess energy between the body and the work environment (Haddon, 1963, 1968, 1980).

The pioneering work of William Haddon during the mid-1900s hypothesized that engineering modifications would make the largest impact and achieve the greatest long-term reductions in injuries. Haddon became the first director of the National Highway Safety Bureau (which later became the National Highway Traffic Safety Administration). He was the impetus behind automotive safety devices like airbags, head rests, collapsible steering wheels, and padded dashboards that lead to an estimated 28,000 lives saved (Guarnieri, 1992). Most relevant to occupational safety and health, Haddon’s engineering philosophies helped develop personal protective equipment (e. g. , ear plugs, hard hats, gloves, safety glasses, steel-tipped boots) for occupational safety.

Throughout the next century, engineers made great contributions to safety by designing safer machinery, using better quality materials, and advancing the design of personal protective equipment the employee would have to endure while operating hazardous equipment (cf. Casali, 1990). However, many engineers tried to completely eliminate the human element by automating many hazardous jobs. Some engineers considered workers “irredeemably flawed and so to be removed from the process” (Hale & Glendon, 1987, p. 2). With an estimated 88% of all industrial accidents being attributed to the at-risk behavior of the employee, this assumption may seem well-founded (Heinrich, 1931). However, a subdiscipline of engineering, human factors (termed ergonomics in Improving Intervention Effectiveness 10 he UK), considers the interaction between the human, machine, and their work environment to be of the greatest importance in causing and preventing injury (Kroemer, & Grandjean, 1997; Roughton, 1996; Tayyari & Smith, 1997). Since the World War II, the discipline of human factors/ergonomics has been gaining prominence in the U. S. as a way to reduce occupational injuries in numerous settings. Ergonomic (or the natural law of work) is the study of the human-machine interface or the science of designing the workplace to fit the employee, and thus reducing the potential for excess energy exchange between the person and their work environment. Guidelines developed by human factors research were originally concerned with increasing human performance or efficiency, with an increase in safety considered a byproduct (Grether, 1975).

Current human factors research focuses not only on productivity, but also on cumulative trauma disorders (including back strain), work spaces, workload, workplace layout, automation, and other physical factors (including temperature, noise, vibration, illumination) that affect workers’ safety and health (Haddon, 1968; Kroemer, & Grandjean, 1997; National Committee for Injury Prevention and Control, 1989; Office of Technology Assessment, 1985; Roughton, 1996; Tayyari & Smith, 1997). Typical human factors programs begin with developing a task force of representatives from various levels and specialties in an organization (e. g. , management, supervisor, engineer, safety professional, employee).

Then, this team or committee develops an occupational hazard survey or some other type of specific environmental audit for assessing the interaction between the employee and his or her work environment (Geller, 1998c; Guastello, 1993). Once the audits are complete, specific recommendations from ergonomic/human factor resources are used for Improving Intervention Effectiveness 11 reducing or eliminating the hazard or at-risk behavior (Guastello, 1993; Kroemer, & Grandjean, 1997; Pheasant, 1991; Roughton, 1996; Tayyari & Smith, 1997). As an occupational safety intervention, human factors programs have been quite successful, with one reviewer estimating an average reduction of 52% in occupational injuries (Guastello, 1993).

However, human behavior plays a major role in every safety-related process. Heinrich, et al. (1980) estimate that 98% of all injuries are preventable, with 88% caused by atrisk behaviors of employees and 10% originating from the hazardous mechanical or physical conditions of the workplace. Thus, human factors programs and engineering modifications have made an impact on the 10% of the hazards in the workplace. However, the remaining 88% of the at-risk behaviors are not being addressed. In other words, engineering and human factors interventions can produce a safer workplace, but “it is difficult to provide a safe work environment solely through safety engineering” (Hoyos & Ruppert, 1995, p. 107).

EDUCATION/TRAINING A common method (or reflexive action) to encourage safe work-related behavior is for organizations to create or purchase an education and/or training program (Jewell, 1998; McAfee & Winn, 1989). In one survey, the majority of organizations (96%) responding indicated that they offered safety training, while another questionnaire found that 46% provided some form of safety training as part of their regular occupational safety efforts (Lee, 1987; McAfee & Winn, 1989). Furthermore, a 1996 survey of over 1200 readers of Industrial Safety and Hygiene News revealed industrial education and training in safety to be a top priority for 1997 (Johnson, 1996).

Educational safety programs focus on increasing peoples’ knowledge by giving them a background on theories, principles and techniques for improving their future problem-solving abilities. In the Psychology of Safety, Geller (1996) stresses the need for safety-related processes Improving Intervention Effectiveness 12 to begin with theory and build from solid psychological principles. Geller also emphasizes the importance of training. Training compliments education by providing employees opportunities to apply the knowledge provided by the education. Thus, the purpose of an education/training procedure “is to provide an environment for the acquisition of attitudes, knowledge or skills, so that newly acquired behaviors may be transferred to the job setting” (Goldstein, 1975, p. 97).

A successful education/training program can impact workers’ safety by giving them the tools and knowledge to use when faced with a novel emergency on or off the job. Viscusi (1983) hypothesizes an alternate motivation for occupational safety and health education/training. In his book, Risk by Choice, Viscusi (1983) examines the motivation behind adopting various types of education/training programs and criticizes the content of such programs. Although employers never provide prospective employees the average annual death risk or chance of acquiring an injury, when workers begin a job they have some general idea of the risks they face. However, once they gain experience on the job, their risk perception changes.

From a sample of 6,000 employees, Viscusi (1983) found that when workers’ risk perceptions increase, their propensity to quit also increases by 35%. Since hiring new employees is costly (due to retraining and loss of experience), the content of education/training programs “is not intended to enable workers to assess the risk more accurately…it is directed at lowering workers’ assessment of the risk” (Viscusi, 1983, p. 71). Consequently, the information given to employees in education/training sessions reduces the perceived risk of their job and avoids costly turnover. Furthermore, results of education/training efforts have been inconclusive, since intervention research seldom solely relies on education/training alone. (Petersen, 1996; Hale & Glendon, 1987).

Petersen (1996) states that “safety training historically has involved more Improving Intervention Effectiveness 13 preachments than real teaching of skills to achieve results” (p. 12). However, Fiedler, Bell, Chemers, and Patrick (1984) found significant improvement in lost-time injuries and number of Mining Safety and Health Administration (MSHA) citations received following a management education/training safety program. Moreover, Zohar (1980) surveyed 20 industries (four from metal fabrication, food processing, chemical, and textile) and found that perceived safety importance was significantly related to the organizational safety climate (i. e. workers’ perceptions of structure, system, goal direction, and management leadership style). Clearly, more research is needed to ascertain the effectiveness of education/training for reducing industrial injuries. Although widespread, education/training programs are rarely systematically evaluated in any type of industrial application (Goldstein, 1975; McAfee & Winn, 1989; Vojtecky & Schmitz, 1986). Evaluative research in occupational safety is seldom initiated because of methodological and design constraints that prohibit many types of teaching methods from being evaluated (U. S. Department of Health and Human Services and Public Health Services, 1990).

For instance, in a survey given to individuals involved in the evaluation of safety and health programs (n = 124), 40% of these safety professionals were dissatisfied with their attempts to demonstrate the beneficial impact of their safety education/training (Vojtecky & Schmitz, 1986). Due to inconclusive findings, occupational safety research needs to address the “longerterm benefit of educational/training programs and how these approaches can be combined with others to accelerate behavior change” (Institute of Medicine, 1988, p. 11). Additional research is also required to identify the conditions under which employees are most likely to participate willingly in the development and implementation of methods to increase the occurrence of safe work behaviors among themselves and others. Finally, if education and training methodologies

Improving Intervention Effectiveness 14 are combined (Geller, 1996), implemented in good faith (Viscusi, 1983), and evaluated systematically to assess the transfer of knowledge (Goldstein, 1975), education/training programs have great potential to make a difference in the safety and health of many employees. ENFORCEMENT The final E in the three Es of safety is enforcement. There are two types of enforcement in occupational safety and health: enforcement within the industry (referred to as discipline) and enforcement by governmental agencies (referred to as compliance). Within industry, the company imposes safety rules or policy and procedures as guidelines for employees to follow.

When employees do not follow these guidelines, there is a possibility of disciplinary action (e. g. , verbal warning, written warning, time off work, job termination). Governmental agencies establish laws or regulations for organizations to follow. When employers do not comply, they receive citations with accompanying fines. For occupational safety and health, it is very common for employees and employers to be held accountable for their actions (Geller, 1998b). There is considerable debate, however, regarding the effectiveness of discipline and compliance as a motivating intervention for safe behavior (Hale & Glendon, 1987; Geller, 1998b, 1996; Petersen, 1996; Wilson, 1985).

Discipline One of the most common techniques used to reduce at-risk behavior within the workplace is to introduce stricter rules, increase supervision of the target behavior or increase the number of reprimands given out for failure to comply with the companies’ policy and procedures (Geller, 1996, 1998b; Hale & Glendon, 1987; Harms-Ringdahl, 1993; McSween, 1995). The introduction of new safety rules followed by discipline for not following those rules can be an effective intervention if delivered correctly. However, discipline is seldom implemented correctly and has Improving Intervention Effectiveness 15 several other negative effects (e. g. , escape behavior, aggression, apathy, countercontrol). In an industrial setting, there are even more barriers in administering discipline (or punishment) in an effective manner.

To have the greatest impact, discipline needs to be given in close temporal proximity to the at-risk behavior. It also needs to be given every time the at-risk behavior occurs. Also, the negative consequences should be sizable (as in severe and aversive) to the employee (Azrin & Holz, 1966; Geller, 1996; Harms-Ringdahl, 1993; Hineline, 1984; Kasdin & Wilson, 1978; McIntire & White, 1975). In the workplace, it is very difficult to give discipline in a soon, certain, and sizable manner. The threat of discipline can suppress behavior, but typically only while the supervisor is observing the employee or until the disciplinary “phase” passes. Pirani and Reynolds (1976) compared the effectiveness of common safety interventions (i. e. safety posters, safety films, fear posters, discussion, role-playing, and discipline) on the use of personal protective equipment (i. e. , hard hat, gloves, safety glasses, and proper footwear). Table 2 was adapted from the results of the Pirani and Reynolds (1976) study where the intervention (Int) percentages represent the percent change over the baseline period, followed by percent change from baseline to a four-month withdrawal (Wdr) period. Discipline did show a moderate increase across all types of personal protective equipment (+39%), but achieved the worst long-term effects, falling an average of 7% below the original baseline periods.

Whereas discipline can be an effective means of altering employees’ safety behavior, it does not seem probable that it can be carried out in an effective manner and may even, in some cases, decrease compliance following its removal (Baldwin, 1992; Geller, 1996; Pirani & Reynolds, 1976). _________________________________ Insert Table 2 About Here _________________________________ Improving Intervention Effectiveness 16 Compliance Organizations encounter many outside safety-related influences that motivate their safety interventions (e. g. , government agencies, union, and professional societies). Due to space, this section will only briefly cover one of the major governmental interventions that use compliance as a motivational technique – Occupational Safety and Health Administration.

During the 1960s, there was a rapid increase in recorded industrial injuries (referred to as delayed progress in Table 1), and by 1967, President Johnson proposed the Occupational Safety and Health (OSH) Act. The OSH Act was finally approved by 1970 with the “purpose of assuring safe and healthy working conditions” (Pease, 1985, p. xiv). This Act created two major institutions that have had a substantial impact on how occupational safety and health interventions are implemented. The National Institute for Occupational Safety and Health (NIOSH) was established to research possible hazards in industry and suggest methods for reducing exposure to these hazards. And OSHA, was created within the Department of Labor to enact regulations and standards to ensure the safety and health of all employees (Viscusi, 1983).

Beyond imposing an average of 45 new regulations/standards every six months (Petersen, 1996), OSHA investigates fatalities and employee complaints, as well as regional programmed inspections. During inspections, the OSHA representative determines if the company is complying with the current standards. If the company fails to meet these standards, they may be fined between $300 and $1,000 with a $10,000 fine (and $1,000 per day until the condition is under compliance) for repeated or willful violations (Petersen, 1996; Viscusi, 1983). Under the General Duty Clause, however, fines from OSHA can range anywhere from $3 to $5 million dollars (Petersen, 1996). On average, OSHA inspectors conduct 60 thousand announced Improving Intervention Effectiveness 17 inspections per year.

With an average of 4 million workplaces, OSHA inspectors can only sample approximately 1. 5% of all industries per year (Pease, 1985). With the inspection being announced, relatively low cost for violations and the extremely low probability of being randomly inspected, the OSHA inspection does not appear to be much of a motivational impetus for occupational safety (Viscusi, 1983). This is not to say OSHA has not had an impact on occupational safety and health. In fact, many companies have placed their whole safety emphasis on the physical aspects the workplace (Petersen, 1996), with the major emphasis placed on performing safety for OSHA, rather than for the safety and health of their employees (Geller, 1996).

Whereas OSHA may affect a companies’ safety professional, it has little impact on the day-to-day interventions used to impact safety in the typical workplace (for a critical review of OSHA and its impact, see Oi, 1975 or Viscusi, 1983). PSYCHOLOGICAL MODELS OF OCCUPATIONAL SAFETY & HEALTH The diversity, intensity, and scope of safety and health education and promotion are immense. There are many target areas (from reducing cigarette smoking to increasing exercise) and governmental objectives (from lowering infant mortality rates to getting 85% of all drivers to buckle-up by 2000) that health professionals or governmental agencies focus on to improve the quality of life for individuals. Some of these areas receive relatively broad study, funding, and promotion (such as reducing drug use among children).

Other areas have a smaller, more specific scope (like getting Amish farmers to place reflective triangles on their wagons). Some of these problem areas and issues are guided by nationally set objectives, as outlined in Healthy People (U. S. Department of Health, Education, and Welfare, 1979) and Healthy People 2000 (U. S. Department of Health and Human Services, 1990). Currently, Healthy People 2000 has 22 priority areas, 300 health objectives, and three overarching aims: 1) Improving Intervention Effectiveness 18 increase the span of healthy life, 2) decrease the incongruity in health between distinct populations, and 3) provide universal access to preventive services.

In accordance with the Healthy People 2000 objectives, health psychologists have made significant progress in achieving the proposed goals by expanding the impact of their interventions across many areas (including home, school, community, work), by using interdisciplinary teams (including sociology, anthropology, communication, marketing, medicine) to synthesize diverse research findings and by developing comprehensive programs to apply results for large-scale behavior change. Health psychologists also study numerous health issues (including smoking cessation, exercise, nutrition, weight control, stress), and they investigate the most effective ways of bringing about change in people’s health-related behaviors (Bennett, Weinman, & Spurgeon, 1990; Elder, Geller, Hovell, & Mayer, 1994; Glanz, Lewis, & Rimer, 1997; Resnick & Rozensky, 1996; Winett, 1998; Winett, King, & Altman, 1989).

One specific area of health psychology that addresses issues in occupational health, is a wellness/worksite health program (Cataldo & Coates, 1986; Kerr, Griffiths, & Cox, 1996; McPartland, 1993; Opatz, 1994). OCCUPATIONAL SAFETY AND HEALTH VERSES WORKSITE HEALTH PROMOTION Within the industrial safety literature, there are research lines drawn between occupational safety, occupational health, and worksite health (Baker, Israel, & Schurman, 1996; Heinrich, et al. , 1980; Opatz, 1994; Tayyari & Smith, 1997). In fact, there is some debate over whether research in industry should be referred to as occupational health and safety (see Baker, et al. , 1996; Goldenhar & Schulte, 1994) or occupational safety and health (e. g. , NIOSH, OSHA).

Occupational safety research looks at injury prevention, engineering/human factors, Improving Intervention Effectiveness 19 education/training, discipline/compliance, and property damage (for a review of property damage see Bird & Germain, 1997). Occupational Health Occupational health focuses on controlling employees’ exposure to occupational disease (e. g. , black lung, TNT poisoning, phossy jaw), while worksite health programs concentrate on individuals’ lifestyles and health-related behaviors (or habits) that may occur on or off the job (Heinrich, et al. , 1980; Kerr, et al. , 1996; McPartland, 1993; Opatz, 1994). See Table 3 for the nine most common worksite health promotion activities.

In terms of safety-related interventions within the workplace, there is considerable overlap of effort between occupational safety, occupational health, and worksite health promotion. Occupational safety and health and worksite health promotion all focus on health behavior, but there is little theoretical overlap in terms of intervention research. _________________________________ Insert Table 3 About Here _________________________________ Health behavior refers to the behaviors of individuals, groups, organizations, communities, and institutions and how those behaviors relate to staying healthy and safe, seeking help when an illness is perceived, and following the appropriate medical advise when sick (Glanz, et al. , 1997; Gochman, 1988, 1997; Winett, 1998).

Specifically, Gochman (1997) defines health behavior as: those personal attributes such as beliefs, expectations, motives, values, perceptions, and other cognitive elements; personality characteristics, including affective and emotional states and traits; and overt behavior patterns, actions and Improving Intervention Effectiveness 20 habits that relate to health maintenance, to health restoration and to health improvement. (p. 3) There are three categories of health behavior: preventive health behavior, illness behavior, and sick-role behavior. Kasl and Cobb (1966a, 1966b) define preventive health behavior as any proactive response taken to maintain a healthy lifestyle (e. g. , buckling safety belts, using PPE, following policy and procedures, see Geller, 1996, 1998a). The other categories (i. e. , illness behavior and sick-role behavior) focus on individuals when they have already been hurt or injured.

Since the targets for most occupational safety and health interventions are proactive or primary in nature (referred to as timing in Winett, 1995), the definition of preventive health behaviors obviously overlaps with the targets of occupational safety and health. Since there is an overlap between occupational safety and health and worksite health promotion, it is surprising there are few health psychologist who expand their intervention focus to occupational safety and health and vise versa (Baker, et al. , 1996; DeJoy, 1996). A prime objective of this paper is to tie together occupational safety and health and worksite health promotion into a model for increasing the effectiveness of interventions to improve both occupational safety and health. Thus, from this point on, behaviors targeted by occupational safety and health interventions and worksite health promotions (i. e. any proactive behavior taken to maintain a healthy lifestyle) are collectively referred to as health behavior (excluding the engineering/human factors research of occupational safety). Occupational Safety Applied behavior analysis has made substantial contributions to the field of occupational safety by documenting the determinants of at-risk behaviors, directing the development of effective behavior change interventions and applying these interventions in a variety of domains, Improving Intervention Effectiveness 21 including behavioral medicine (Cataldo, & Coates, 1986), safety management performance (Daniels, 1989), health behavior (Elder, et al. 1994), traffic safety (Geller, 1998a), behaviorbased safety (Geller, 1998f; Krause, Hidley, & Hodson, 1996), environmental protection (Geller, Winett, & Everett, 1982), safety management (Petersen, 1996), child safety (Roberts, Fanurik, & Layfield, 1987), and health psychology (Winett, et al. , 1989). Applied behavior analysis focuses on behavior and “what people do influences the quality of life, and people doing is the realm of psychology, the science of behavior” (Roberts, et al. , 1987, p. 105). In occupational safety and health, there has been a strong application of applied behavior analysis principles referred to as behavior-based safety (BBS). For the past twenty years, BBS has been used successfully in the prevention of occupational injuries (e. g. Alavosius & SulzerAzaroff, 1986; Geller, Davis, & Spicer, 1983; Geller, Eason, Philips, & Pierson, 1980; Geller, Roberts & Gilmore, 1995; Komaki, Barwick, & Scott, 1978; Komaki, Heinzmann, and Lawson, 1980; Pettinger, Boyce, & Geller, in press; Reber & Wallin, 1983, 1984; Roberts & Geller, 1995; Reber, Wallin, & Chokar, 1990; Smith, Anger, & Ulsan, 1978; Streff, Kalsher, & Geller, 1993; Williams & Geller, in press). Table 4 summarizes the findings of a review of 53 occupational safety and health interventions since 1977. In his systematic review, Guastello (1993) discovered that BBS has the highest average reduction (59. 6%) of injury rate (see Table 4) compared to other occupational safety intervention strategies. _________________________________ Insert Table 4 About Here _________________________________ Improving Intervention Effectiveness 22

Although BBS approaches differ in various ways, they all focus on systematically studying the effects of employees using an observation and feedback process to assess their organization’s safe and at-risk work related behaviors. Using this information, employees then design interventions to target critical at-risk behaviors and decrease their occurrence. Typically, BBS techniques accomplish this by first having the employees define target behaviors in an observable, recordable and trackable manner, and then observe and record them in a typical work setting. Following a relatively stable baseline measure of the frequency, rate, or duration of a target behavior, employees introduce an intervention to change the targeted behavior. Interventions typically involve adding or changing antecedents and/or consequences of the target behavior(s).

To determine intervention effectiveness, baseline frequency, duration, or rate of the target behavior is compared to post-intervention measures (e. g. , Daniels, 1989; Geller, 1996, 1998f; Komaki, & Jensen, 1986; McIntire, & White, 1975; McSween, 1995; Raouf, & Dhillon, 1994). The BBS approaches to occupational safety and health have reported a number of advantages over other industrial safety techniques, including the ability: 1) to be administered by employees with a minimal professional safety background, 2) to reach people in the setting where a safety problems occur, and 3) to be readily customized for application in a specific safety-related work culture (Daniels, 1989; Geller, 1998e, 1998f; McSween, 1995; Roberts & Geller, 1994).

Previous research has also shown the BBS approach to be cost effective, relatively easy to administer, and straightforward enough to be assessed by employees who track the target Improving Intervention Effectiveness 23 behaviors (e. g. , Daniels, 1989; Geller, 1996, 1998f; Pettinger, et al. , 1999; Sulzer-Azaroff & De Santamaria, 1980; Sulzer-Azaroff, Loafman, Merante, & Hlavacek, 1990). Another advantage claimed by Geller is when delivered correctly, BBS can motivate employees to actively care for their fellow coworkers. Actively caring is described by Geller (1991) as going beyond the call of duty for the safety and health of others by taking personal action to correct hazardous conditions, increase safe behavior, or decrease at-risk behavior (Geller, 1996; Geller, Roberts, & Gilmore, 1996; Roberts & Geller, 1995).

Consequently, BBS has become such a force in occupational safety and health that a variety of other safety disciplines are supplementing their approaches with behavior analysis techniques. For example, many safety management techniques now incorporate behavioral observations (Pybus, 1996; Yandrick, 1996), or combine successful management techniques like “Systems Management” (Waring, 1996), “Total Quality Management” (Weinstein, 1997), and “Management by Objectives” (Petersen, 1996) with a BBS focus. Actively Caring Actively caring is a conceptual approach to injury prevention which focuses on identifying individuals who are most likely to take personal action to correct unsafe conditions, ncrease safe behavior, or decrease at-risk behavior (Geller, 1991, Geller, 1996; Geller, et al. , 1996; Roberts & Geller, 1995). In other words, the model of Actively Caring refers to going beyond the call of duty for the safety and health of others as opposed to simply caring for the safety of oneself. Thus, by increasing actively caring behaviors, one may affect the culture and in turn influence more people to actively care. The Actively Caring model has been used successfully in industrial settings to predict employees’ willingness to actively care for a Improving Intervention Effectiveness 24 coworker’s safety (Geller, 1996), and predict who will participate in a safety recognition process (Roberts & Geller, 1995).

Behavior-based safety interventions attempt to motivate employees to observe each other and provide proactive feedback regarding safety, which can be interpreted as “actively caring”. Consequently, assessing whether employees feel confident in their abilities to perform the tasks required of them in a BBS process should be important in determining if they will participate in a safety process like BBS. Additionally, if employees actively care on a regular basis and a BBS process is introduced into their organization, those individuals who already look out for safety and health of others (or actively care) would be predicted to participate more than employees who do not typically look out for the safety or health of coworkers.

Therefore, from an actively caring perspective, it’s important to assess not only how confident employees feel in performing actively caring behaviors, but also to determine their current state of self-esteem, self-efficacy, personal control, optimism, and belonging, factors presumed to be directly related to amount of actively caring (Geller, 1996; Geller, et al. , 1995; Roberts & Geller, 1995). MODELS OF SAFETY & HEALTH Along with the diversity of safety and health promotion and intervention approaches, many different theories and models address safety and health behavior. In fact, many of the approaches to safety and health intervention are derived from some theory or model of behavior change. In fact, Glanz, et al. 1997) surveyed 24 health behavior journals from 1992 through 1994 and found that out of 1,174 relevant articles, 45% focused on at least one theory or model. In this review, the authors found 66 different theories or models in use. Table 5 provides a list of the 20 most cited theories/models related to health behavior. _________________________________ Improving Intervention Effectiveness 25 Insert Table 5 About Here _________________________________ From 1990 to 1999, there have been at least 11 publications that proposed or adapted models for improving health behaviors. Many of the models have targeted specific health-related behaviors like increasing earplug use (Lusk, Ronis, & Kerr, 1995), reducing overexertion (Kumar, 1994), or controlling occupational stress (Baker, et al. , 1996).

Three models focus more on environmental factors that influence safety such as safety management (DeJoy, 1994) and broad organizational factors leading to occupational injuries (DeJoy, 1990; Reason, 1995). Several models focused more on presenting a framework or stage process to achieve the desired health behavior (Andersson, & Menckel, 1995; DeJoy, 1996; DeJoy & Southern, 1993; Geller, 1998d; Geller et al. , 1990). With few exceptions, the models mentioned above fail to provide the safety practitioner with guidelines for increasing the beneficial impact of safety-related intervention. Some authors do suggest content for interventions such as self-efficacy training, supervisor training, videotape training. But most models only provide a conceptual basis for the design and/or implementation of a behavior-change program (Conner & Norman, 1996).

In a comprehensive review of occupational safety and health intervention, Goldenhar and Schulte (1994) conclude that research on the application “of etiological knowledge and on ways to conduct occupational safety and health intervention research is sparse” (p. 763). THEORETICAL BASES FOR PROPOSED RESEARCH This research compares theoretical constructs from social cognitive theory and other health behavior frameworks. More specifically, this research 1) assessed the potential contributions to health behavior intervention of self-efficacy/outcome expectancy (Bandura, 1997) and the stages of change (Prochaska, et al. , 1997), and 2) compared the ability of self- Improving Intervention Effectiveness 26 fficacy/outcome expectancy versus stages of change to predict involvement in a behavior-based safety intervention process. SOCIAL COGNITIVE THEORY The concept of self-efficacy and outcome expectancy has been widely used in a variety of health-related settings (Bandura, 1997; Maddux, 1995). Even though there are many health related models that have had been used successfully to design interventions to achieve positive results (e. g. , health belief model, theory of reasoned action, theory of planned behavior, protection motivation), self-efficacy was chosen because of the many successful applications of the theory in a variety of settings and also due to significant overlap of determinants between social cognitive theory and similar health related theories.

Table 6 compares determinants of health behavior by listing five commonly used models, and depicts the significant overlap between social cognitive theory and the others listed. _________________________________ Insert Table 6 About Here _________________________________ In a history of social cognitive psychology, Barone, Maddux, and Snyder (1997) traced the American forerunners of the sub-field to such early psychologists as John Dewey (18591952), James Baldwin (1861-1934), and George Mead (1863-1931). Although Dewey, Baldwin, and Mead were discussing such things as philosophy, developmental psychology, and sociology, these early psychologists laid the groundwork for two social cognitive theory tenets still used today.

The first tenet describes how psychology needs to include the social context within the study of human behavior because people are essentially social in nature. The second principle illustrates how we use our cognitions for avenues of thinking and communicating to adapt to Improving Intervention Effectiveness 27 social contexts. Thus, “social cognitive psychology construes cognition as a part of social acts” (Barone, et al. , 1997, p. 11). Of the early frames of reference for social cognitive theory (i. e. , social gestalt, constructivist, information processing, social learning), this section focuses on selfevaluation/regulation (i. e. , control theory, goal-setting theory, self-efficacy). This section is further narrowed to self-efficacy theory Bandura, 1977, 1986, 1997) in that goal-setting theory (Lock & Latham, 1990) overlaps with self-efficacy theory, and control theory (Carver & Scheier, 1981; Scheier & Carver, 1988) is seen as too limiting or mechanistic to cover the diversity of human behavior (Barone, et al. , 1997). Throughout life, people strive to gain control of the various aspects of their environment. Individuals try to gain control over desired outcomes (or attainments) and achieve control over the undesirable events. From a social cognitive perspective (cf. Bandura 1986; 1997), people are exposed to various interdependent circumstances every day (i. e. , reciprocal causation), determine the best approach to these situations, assess their perceived competence (i. e. , self-efficacy) to carry out their intentions (i. e. human agency), determine if the behavior they perform will produce the desired outcome (i. e. , outcome expectancy), and finally decide the importance of obtaining the outcome (i. e. , outcome value). Self-Efficacy Self-efficacy, originally defined as a person’s belief in his or her ability to perform a specific behavior to produce an outcome (Bandura, 1977), has since been expanded by Bandura (1997) to refer to “beliefs in one’s capabilities to organize and execute the courses of action required to produce given attainment” (p. 3). Efficacy beliefs can vary in level (increasing difficulty of behavior), generality (similarity of behaviors), and strength (perseverance). Improving Intervention Effectiveness 28

From Bandura’s perspective (1997; also see Maddux, 1995), people’s self-efficacy influences many aspects of their every-day life. Once an individual’s self-efficacy forms for a particular behavior or set of behaviors, these beliefs guide the person’s aspirations, behaviors, efforts, and reactions. However, these behaviors are seen more as probabilistic rather than an inevitability through reciprocal determinism (Bandura, 1986; also see Geller, 1996, 1998e, 1998f). In other words, three interdependent factors, behavior, person, and environment, influence each other depending upon the situation. Outcome Expectancies Outcome expectancies have also played an important role in social cognitive theory.

An outcome expectancy is the belief that a particular behavior will result in a certain outcome. Outcome expectations take three different forms: physical, social, and self-evaluative. Physical outcomes of engaging in a behavior can be pleasant or aversive sensory experiences. There can also be positive social outcomes such as interest, praise, and recognition, as well as negative social outcomes like disapproval, rejection, or penalties. People also have certain outcome expectancies about how they view themselves. Whereas outcome expectancies have different forms (i. e. , physical, social, and self-evaluative), all of these forms can vary in their importance or value.

Outcome value (Maddux, 1995; Maddux, Norton, & Stoltenberg, 1986) has been recently proposed as another significant predictor variable within self-efficacy theory with its own moderators. Outcome value can vary in dimension, displacement, and velocity. Taylor (1991) proposed that an outcome’s dimension, positive or negative properties, could differentially affect people’s emotions or moods. For example, negative outcomes could produce more cognition, affect, physiology, and behavior in some people than the opposite positive outcomes. Improving Intervention Effectiveness 29 Furthermore, satisfaction with outcomes varies with their displacement from outcome value expectancies to post outcome change (i. e. , displacement relation). In addition to displacement relation, Hsee and Abelson (1991) also proposed velocity relation.

The authors found that people are not only engaging in a behavior to receive an outcome, they are seeking a greater rate (or velocity) of change in the outcome itself, except if it’s negative. Safety Education and Self-Efficacy/Outcome Expectancy Safety-related education occurs in industrial settings almost reflexively (Petersen, 1996). From a social cognitive standpoint this can have a variety of effects. For instance, typical safety education sessions focus either on giving employees information regarding hazardous conditions or use scare tactics to warn employees about dangerous safety-related situations. According to Bandura, neither method would alter employees’ self-efficacy.

To have the greatest impact on employees’ self-efficacy (for specific safety-related topics), “a shift in emphasis is required, from trying to scare people into health to providing them with the tools needed to exercise personal control over their health habits” (Bandura, 1997, p. 280). Therefore, to have an effect on employees’ safety self-efficacy, safety education needs to focus on providing employees with training to give them the needed skills to perform their work tasks safely. While a typical safety education session might not impact employees’ self-efficacy, it could influence their outcome expectancies. If employees watch a safety video which depicts a finger amputation occurring because an employee failed to turn off the power to a machine (e. g. lock-out/tag-out), a shift in the viewers’ outcome expectancies could change toward that particular type of injury. The severity of the injury would produce expectancies concerning the physical disability (i. e. , a physical outcome expectancy) from the amputation, aversive social Improving Intervention Effectiveness 30 reactions from family, friends, and coworkers (i. e. , a social outcome expectancy), and if the person held safety as a core value, a negative self-evaluation. In summary, from a social cognitive view, the combination of safety training and safety education could increase employees’ self-efficacy and shape their outcome expectancies if they had quality training and believe they can have input in the safety process.

Consequently, safety interventions that focus on providing practical tools and methods for improving safety should increase participants’ safety self-efficacy regarding injury prevention. SEGMENTATION TECHNIQUES TO BEHAVIOR CHANGE Transtheoretical Model of Change The transtheoretical model was developed out of a comparative analysis of more than three hundred studies of psychotherapy and behavior change (Prochaska, 1979). The analysis identified ten techniques or “processes of change” used in attaining particular health behaviors. Next, DiClemente and Prochaska (1982) compared two groups of individuals (self-treatment vs. professional treatment) trying to quit smoking. The authors assessed how often each group used a specific process of change.

From interviewing subjects, Prochaska and DiClemente (1983) discovered that individuals trying to quit smoking used different techniques or “processes” depending upon where they were in the quitting process. Table 7 below lists the stages of change and the ten processes of change with brief descriptions of each. _________________________________ Insert Table 7 About Here _________________________________ Prochaska, et al. reported (1997) that on average 40% of current smokers (or other risky health behaviors) were not even considering quitting (i. e. , precontemplation), 40% were at least Improving Intervention Effectiveness 31 thinking about quitting (i. e. contemplation), and 20% had an actual plan to quit (i. e. , preparation). Prochaska and DiClemente then argue that the typical smoking programs, which are “action-oriented” interventions would not be appropriate for individuals in the Precontemplation phase. Rather, health practitioners need to tailor their interventions to the specific stages of their targeted population. Presumably, people performing a risky behavior, go through four stages of change on their way to acquiring and maintaining a safe behavior: precontemplation, contemplation, preparation, action, and maintenance. The stage of change model, has been criticized by Bandura (1997, p. 13) as a “descriptive device” that “provides no explanation for why [people] do not consider making changes” (cf. Davidson, 1992). Smedslund (1997), however, argues that not all psychological research must have an theoretical base to be useful for producing beneficial behavior change. Although the stages of change may not be considered empirical, the “model can, however, be used to generate empirical research” (Smedslund, 1997, p. 542). Prochaska, DiClemente, Velicer, and Rossi, (1992) address these concerns directly by arguing that the transtheoretical model’s primary purpose is in achieving behavior change, not studying behavior acquisition (see a review in Health Education Quarterly Vol. 9(3), 1992). DiClemente and Prochaska (1985; DiClemente et al. , 1991; Prochaska, DiClemente, & Norcross, 1992) have examined self-efficacy’s role in progressing through the stages of change and found that for addictive behaviors, individuals’ self-efficacy was significantly related to their stages of change. Furthermore, DiClemente, Fairhurst, & Piotrowski (1995) suggest that selfefficacy enhancing strategies should be used for the latter stages (i. e. , action and maintenance), while interventions addressing outcome expectancies for motivating decision making are better suited for the early stages of change (precontemplation, contemplation).

Improving Intervention Effectiveness 32 Whereas the stages of change model may be viewed by some as only a descriptive tool that does not consider each employee’s individual determinants for not performing a particular safe behavior, numerous studies have demonstrated the positive impact of customizing interventions to a recipient’s stage of change (DiClemente, et al. , 1995; Prochaska, et al. , 1997). Although there have been many applications of the transtheoretical model across numerous types of health behaviors (involving smoking, condom use, exercise, weight control, diet, sunscreen, mammography), there is no application to occupational safety.

Some worksite health programs may have used this technique, but none have focused their efforts on safetyrelated at-risk behaviors. If employees are engaging in at-risk safety-related behaviors, the stages of change can be applied to assess which stage these individuals are at regarding their motivation to change. Thus, by segmenting an employee population in terms of stages of change, the safety leaders in an organization can customize an intervention strategy to target specific safety education/training for individuals at varying levels of change. Since there has been no direct application of the stages of change or the efficacy of the 10 processes of change to occupation safety, additional models of industrial safety intervention need to be consulted.

Safety Stages of Change One industrial safety model that uses a segmentation methodology similar to the stages of change and suggests industrial safety intervention strategies is the flow of behavior change model. The flow of behavior change model originally conceived by Geller (1998d), assists occupational safety and health professionals design interventions to fit the problem. Geller hypothesized that employees fall into one of four stages with regard to their type of intervention needed. Once an organization’s leadership segments their employees into these stages, Geller suggests guidelines for developing safety interventions. Improving Intervention Effectiveness 33

More specifically, Geller proposes three types of behavior: other-directed, self-directed, and automatic behavior. When a behavior is first learned, Geller hypothesizes it is other-directed. With other-directed behaviors, individuals rely on external guidance or motivation to become competent in that particular behavior. Once the person practices the newly acquired behavior, it can then become self-directed (cf. Watson & Tharp, 1993). If the behavior is difficult, undesirable, or the response effort is too great (as when the amount of perceived effort involved in performing the behavior outweighs the perceived benefits), the behavior may never become self-directed and will always need some type of extrinsic incentive or disincentive.

Self-directed behaviors are motivated from within people (i. e. , internal). If employees, for example, are wearing steel-toed boots because it is part of the company’s safety policy and would not wear them otherwise, that person is other-directed. On the other hand, if employees wear their steel-toed boots because they perceive themselves safe workers and would wear steeltoed boots regardless of the safety policy, that individual would be considered self-directed. Over a period of time, if a behavior is repeated frequently and consistently, it may become a habit (i. e. , automatic, see Posner & Snyder, 1975; Schneider, & Shiffrin, 1977; Shiffrin, & Schneider, 1977).

Furthermore, if a behavior is not very complex (e. g. , connecting a safety-belt), it can move immediately from other-directed to becoming a habit, or from a risky habit to a safe habit. Pettinger and Geller (1999) expanded Geller’s original flow of behavior change model by modifying Prochaska and DiClemente’s stages of change model for industrial safety application. Pettinger and Geller’s safety stages of change (see Figure 2) hypothesizes that individuals transition through four stages or levels of competence (precontemplation, contemplation, action [other-directed or self-directed], and maintenance) when moving from an at-risk habit to a safe habit.

In the new model, employees are not aware they are doing something risky or they don’t Improving Intervention Effectiveness 34 know the correct safety procedure (i. e. , precontemplation), they know the correct procedure but are still engaging in the at-risk behavior (i. e. , contemplation), they consistently perform the desired safe behavior but still need to think about the correct procedures (i. e. , action), or they no longer have to consciously perform the safe behavior, the individual is at the maintenance stage (i. e. , safe habit/automatic behavior). Table 8 below provides definitions for each of the stages hypothesized in the safety stages of change. ________________________________ Insert Figure 2 About Here _________________________________ Employees can be in one of two different “action” stages for a particular behavior – action self-directed and action other-directed. For example, if employees are performing safe behaviors consistently, but need external direction to maintain the behavior (the perception that an external source is setting the contingencies for that behavior), they are at the action stage (other–directed). If employees are performing the safe behavior consistently and no longer require any external motivation (contingencies are internally set by the individual), they are in the action stage (self-directed). ________________________________ Insert Table 8 About Here _________________________________ The safety stages of change model suggests four types of behavior-change interventions, depending upon the stage of the target performer: instructional, motivational, supportive and self-management. When willing learners are first acquiring new behaviors, teachers typically rely on instructional interventions that use antecedent strategies (Geller, 1996; Geller, 1998d) to Improving Intervention Effectiveness 35 move people from the precontemplation stage to the contemplation stage. Instructional interventions can be any antecedent event that directs behavior.

In industrial settings, education/training sessions often use training videos to teach new procedures or demonstrate methods of handling new safety situations. These sessions are designed to be clear and easy to understand in order to facilitate movement to the contemplation stage. Although some people might understand the information given during the instructional intervention, they may not believe the information to be pertinent or they might receive greater rewards for disregarding the instructions. For example, people may take calculated risks or shortcuts to accomplish a job faster (like changing a light bulb while standing on a chair instead of a ladder).

The model would place this individual in the contemplation stage. To move employees from contemplation (meaning they know what to do but do not do it) to action, the safety stages of change model proposes a motivational intervention, an intervention that uses both antecedent and consequence strategies. Since an employee in the contemplation stage knows the safe behavior, an antecedent strategy alone will have minimal impact (Geller et al. , 1990). Using a consequence strategy along with an antecedent strategy is presumed to be the intervention strategy of choice in order to encourage people to engage in the target safe behavior (Geller, 1996; 1998f).

Geller and his colleagues, for example, have used many incentive/reward strategies (combining activators and consequences) to motivate individuals to use their safety belts (e. g. , Geller et al. , 1990). Once the employee is performing the safe behavior (because of either self- or otherdirections), supportive interventions can help people move from the action stage to the maintenance stage (i. e. , a safe habit). Supportive interventions mostly involve consequence strategies that reinforce and support desired behaviors. Supportive interventions may include Improving Intervention Effectiveness 36 recognition, interpersonal praise, and peer encouragement. Employees in the action self-direction stage derive their reinforcement from within the task itself, thus, safety self-management (cf.

Mahoney, 1971, 1972) is an intervention that empowers employees to set their own contingencies and provide their own reinforcers. In summary, the safety stages of change suggests that intervention impact is greatest if the type of intervention method uses the readiness for change stage of each potential participant. In other words, in order for an intervention to have the greatest impact on the safety-related behaviors of employees and facilitate progress to a safe habit, organizations need to segment their employees into the various stages hypothesized by the safety stages of change and design stage-specific safety improvement processes. For example, if an organization wanted to motivate their employees to perform specific safety-related behaviors through a lantwide educational intervention, the safety stages of change would predict the most behavior change to occur for employees in the Precontemplation or Contemplation stages. RESEARCH OVERVIEW & HYPOTHESES The primary purpose of this research was to compare self-efficacy theory (Bandura, 1997) and the stages of change model (Prochaska, et al. , 1997) as a determinant of safety-related behaviors and amount of involvement in a behavior-based safety process. Additionally, the predictive validity of the actively caring concept combined with the stages of change model, was examined to determine any relationships in participating in a BBS process. H1: From a social cognitive perspective, the BBS intervention will provide employees with the knowledge and tools to perform the safety-related activities taught. H1. : Employees’ safety self-efficacy at the start of the intervention phase will predict involvement in the BBS process. Improving Intervention Effectiveness 37 H1. 2: Employees’ safety outcome expectancy assessed at the start of the intervention phase will predict involvement in the BBS process. H1. 3: Safety perception instruments based on the social cognitive perspective will be the most predictive of employee involvement in a BBS process. H2: From a safety stages of change perspective, the BBS intervention will differentially effect employee involvement based on their current safety stage. H2. 1: Employees’ safety stages of change will predict involvement in the BBS process. H2. : Within the safety stages of change, there will be two factors: a self-directed safety stages of change and an actively caring safety stages of change. H2. 3: The survey designed to place people in certain safety stages of change will be the most predictive in assessing employee involvement in the BBS process. H3: Safety measures will be differentially employees’ safety self-efficacy, safety outcome expectancy, and safety stages of change. H3. 1 An inverse relationship will occur between safety outcome expectancy and injury outcome statistics (i. e. , OSHA recordables, lost time injuries). H3. 2 An inverse correlation will occur between safety self-efficacy and injury outcome statistics. H3. : An inverse relationship will occur between employees’ stage of change and injury statistics. METHOD OVERVIEW To assess the predictive utility of safety self-efficacy and the safety stages of change, the revised Safety Self-Efficacy and Safety Stages of Change surveys were given to two Improving Intervention Effectiveness 38 manufacturing facilities prior to implementing a behavior-based safety (BBS) process. Following the organization-wide education/training, involvement measures, as well as other safety statistics over a four-month period were collected for each employee. PARTICIPANTS & SETTINGS Over three years, ten industrial sites have contributed to the development and testing of the current research.

Preliminary survey development and refinement were conducted with six industrial interview sites and two pilot companies. Following survey development, two additional industrial locations were recruited and served as experimental sites. After administering the revised surveys to the experimental sites, the two organizations received behavior-based safety education/training. Employee involvement in the new safety process was tracked over four months. Injury statistics were acquired from one experimental site. Industrial Interview Sites Preliminary survey work conducted with companies in Milwaukee Wisconsin, Chicago Illinois, Elkin North Carolina, Cowpens South Carolina, York Pennsylvania, and Bakersfield California preceded initiation of this research.

The interview sites represented small to medium sized facilities ranging from 50-500 employees, across a variety of industries including medium to light manufacturing, maintenance, and power generation. Pilot Company-A Participants were 495 employees from a bearing manufacturing facility in the southeast region of the United States. The average age of the workers at the Pilot Company-A is 32 years, 82. 5 percent are white, and 58. 5 percent are male. The average education level of the hourly workers is a high school diploma. Improving Intervention Effectiveness 39 Pilot Company-B Participants were 450 employees from a rotary motor manufacturing facility in the southeast region of the United States. The average age of the workers at the Pilot Company-B is 45 years, 88 percent are white, and 56 percent are male.

The average education level of the hourly workers is a high school diploma plus one year of college/technical school. Experimental Site-NP Participants were 530 employees from a newsprint manufacturing facility (NP) in the northwest region of the United States. The average age of the workers at NP is 40 years, 93 percent are white, and 61 percent are male. The average education level of the hourly workers is a high school diploma plus two years of college/technical school. Experimental Site-TK Participants were 350 employees from a rotary bearing manufacturing facility (TK) in the southeastern region of the United States. The average age of the workers at TK is 37 years, 88 percent are white, and 56 percent are male.

The average education level of the hourly workers is a high school diploma plus one year of college/technical school. DEVELOPING SAFETY PERCEPTION INSTRUMENTS Item Development Over a two-year period, six industrial sites were visited and unstructured, informal interviews scheduled. These 30-minute interviews took place during an eight-hour safety training class with the group size averaging 9 to 30. The interviews were open-ended, group-focused exercises with the companies’ safety administrators, as well as a sample of employees representing various areas of the organization including upper level management, maintenance, operations, engineering, and laboratory technicians.

In all, it was estimated that over 800 Improving Intervention Effectiveness 40 employees and management participated in these initial interviews. These interviews guided the construction of the initial survey items for both the Safety Self-efficacy and Safety Stages of Change instruments. During the unstructured focus groups, the primary researcher asked questions assessing safety impediments and challenges faced by employees on a daily basis like “What is holding you back from being safe? ” and “If you worked in a Total Safety Culture, what would it be like? ” Although the questions were not scripted, they did consistently focus on three safetyrelated factors: environment (e. g. “what are some environmental things that make it difficult to do your job safely? ”), person (e. g. , “what are some situations that make you feel less confident in performing your job safely), behavior (e. g. , what are things people do to put themselves at-risk for an injury) (see Geller, 1996). Such questions assisted the author in determining what would be relevant safety gradations for creating safety self-efficacy questions. Also, by inquiring what percentage of people comply with the company’s safety requirements and asking what are some examples of employees looking out for their coworkers safety (i. e. , actively caring), the author was able to create theoretical “safety stages”.

Safety Self-Efficacy The initial 22-item Safety Self-efficacy questionnaire in Appendix A contains both selfefficacy measures, as well as outcome expectancy items. The self-efficacy questions (n=11) were designed to tap into employees’ Safety self-efficacy level (n=5) and their self-efficacy strength (n=6). People’s efficacy level reflects the number of steps of increasing difficulty they believe themselves capable of performing (i. e. , Even if a safety procedure is long or complicated, I am able to follow it). A person’s efficacy strength assesses persistence in the face of other barriers to performance (i. e. , I avoid trying to learn new safety procedures when they look difficult for me). Improving Intervention Effectiveness 41

The outcome expectancy items (n=11) were designed to assess an employee’s physical (n=3), social (n=4), and self-evaluative (n=4) expectancies. Physical expectancies are any sensory experiences brought about by engaging in a particular behavior (i. e. , I never follow safety procedures because they take too much time). Social expectancies are the expected reactions from the employee’s social systems (e. g. , recognition, status, power, disapproval) resulting from a particular behavior (i. e. , I often skip safety rules when there are large production demands). Self-evaluative expectancies are reactions to one’s own behavior (i. e. , If I skip a safety protocol, I feel disappointed in myself).

Safety Stages of Change The Safety Stages of Change survey was designed to place employees into one of four stages: precontemplation, contemplation, action, or maintenance with regard to their use of personal protective equipment (PPE), lifting techniques and lockout/tagout operating procedures. The safety-related behaviors on the Safety Stages of Change survey were chosen based on the preliminary work at the 6 industrial interview sites. These behaviors focused personal safety related behaviors. Appendix B gives the three questions used to place employees into one of the four safety stages of change for PPE, lifting and lockout/tagout. PROCEDURE The procedures for implementing the BBS process were held constant over the two experimental sites NP and TK.

The only variables that changed were the number of participants and types of critical safety-related behaviors the sites focused on. All workshops were lead by the author and one other graduate research assistant. At both experimental sites, the BBS process was implemented in three steps: a) form a “safety steering team” b) conduct an intensive safety steering team workshop, and b) provide workshops for all employees. Improving Intervention Effectiveness 42 The safety steering teams consisted of volunteers from representative areas of the organization (e. g. , production, maintenance, lab). Experimental site NP had 14 safety steering team members, while site TK had 10 members.

At each site, this committee received three eighthour sessions of BBS education and training to: a) foster the belief that the BBS process is employee driven, b) become the in-house experts of the BBS process, c) discuss which behaviors were critical to the safety of their coworkers, e) design a tool to measure their safety-related behaviors, and e) lead area-specific safety meetings in setting goals, discussing the results of their behavioral observations, and celebrating safety achievements. Following the safety steering team workshops, all employees at both experimental sites (including the committee members) received eight hours of BBS training. Survey Administration Before the safety steering team and employee workshops began, all participants completed both the Safety Self-efficacy and Safety Stages of Change surveys.

The employees were instructed to respond truthfully, and were ensured their responses would be kept strictly confidential. Both experimental sites were concerned with employee anonymity, thus a code number was used to track employees. The code number used at both experimental sites was referred to as an employee’s “clock number. ” The clock number was an already established internal tracking system; therefore no participants’ names were recorded on the surveys. On average, the questionnaires took no longer than 30 minutes to complete with very few questions or comments regarding the survey items. The experimental sites, however, required that completing the surveys be voluntary. Improving Intervention Effectiveness 43

Safety Steering Team Workshop The first two days of the safety steering team workshop focused on the principles, concepts, and critical aspects for implementing and institutionalizing a BBS process. Workshop content and exercises where adapted from Geller 1998e and 1998f. The third day of the workshop focused on site-specific application issues such as what type of work-related behaviors are critical to target, and what the initial critical behavior checklist (CBC) should look like. During the third day of the workshop, the safety steering teams from NP and TK developed a critical behavior checklist (see Appendix C) to be used by their employees following the site-wide employee BBS training.

Once the site specific behaviors were chosen, they were operationally defined to ensure objective information would be collected. For example, “not paying attention while lifting” would be very subjective and open for interpretation, while “did not bend knees when lifting” is much more specific and objective. The CBC is a powerful safetyimprovement tool that provides objective information regarding current site-wide safety performance, presents feedback to workers to guide their future safety-related behaviors, and guides safety improvement activities (Geller, 1996, 1998e; Krause, et al. , 1996). Employee Workshop Following the safety steering team workshop, all employees were given eight hours of BBS training.

The information covered in the employee training classes was based on the materials presented to the safety steering team, but in an abbreviated eight-hour interactive workshop. The principles, concepts, and tools presented to the employees were consistent with those covered in the safety steering team classes, however there were fewer exercises and opportunities for employees to provide input concerning safety process design and implementation. During a 30-minute portion of the employee workshops, one member of the Improving Intervention Effectiveness 44 safety steering team described the process in which they developed the CBC and illustrated how to conduct a safety observation and provide feedback to their fellow coworkers.

Behavior-Based Safety Intervention After each experimental site completed their BBS training, everyone who received training was encouraged to use their CBC to conduct safety behavioral observations and give each other feedback regarding safe versus at-risk behaviors they observed. Observations are initiated by employees approaching fellow coworkers and asking if they are willing to participate in the new safety process and be observed using the CBC. It is then that employee’s decision to participate in the safety observation process. The employee being observed is kept anonymous to counter the perception that the CBC could be used for punitive consequences.

Anonymity increases acceptance of the entire BBS initiative. The name of the observer, however, is recorded to track employee involvement. Following a completed safety observation, the observer places the CBC in a collection box located in various areas around the organization. Once a week, a safety steering committee member collects the observation forms and enters the data into a safety involvement database. Both experimental sites used a database to track which employees were conducting safety observations. In addition to tracking employee involvement, experimental site NP also recorded number of injuries 24 months prior to initiation of the BBS process.

RESULTS SURVEY REFINEMENT Pilot Company-A To further refine the surveys items, the preliminary versions of the Safety Self-efficacy and Safety Stage of Change surveys were administered to employees (N=495) at Pilot Company- Improving Intervention Effectiveness 45 A. The surveys were given out over a one-month period, area by area, during each group’s monthly safety meeting. Results of this pilot research were encouraging. Pilot Company-A provided a 65% return rate for the preliminary versions of the Safety Self-efficacy and the Safety Stages of Change surveys (n=350). Inter-item correlations were performed following the administration of the Safety Selfefficacy survey (n=287). Using alpha as a measure of internal consistency, the scale produced a score of . 88. The sub-scale of