Thomas E. Bernard and Kimberly Y. Anderson

College of Public Health,
University of South Florida, Tampa FL

Abstract

When ergonomic principles are not followed in the workplace, injuries and illnesses along with losses of quality and productivity may occur. The growing recognition of these problems has led to the development and adoption of various means to perform an ergonomic evaluation. Three tools were considered for the analysis of jobs in an automotive assembly plant. Among many, one was proposed by Rodgers in 1985, a second by OSHA in 1995, and a third (Ergonomics Analysis Package) by Bernard in 1995. The Rodgers Tool emphasized the potential for local muscle fatigue by considering the degree of effort, the period of exertion, and the frequency of repetition. OSHA prepared a draft checklist with various scores associated with each item. The Ergonomics Analysis Package was developed to assess the capacity or acceptability with respect to strength limits, fatigue limits, and repetitive motion along with manual materials handling.

Data were collected for 267 different workstations. The data for the Ergonomics Analysis Package was summarized and a Concern Level with values from 1 to 5 was assigned. The overall score for the job was the highest individual Concern Level. The OSHA Checklist was divided into three sections: upper extremities, back, and lifting. The total score was computed for each section based on the values OSHA assigned to different risk factors. The overall score was the highest for the upper extremities section and back plus lifting section. For the Rodgers’ Tool, a Priority Score None, Moderate, High, or Very High was assigned, and the overall score was the highest of the individual priority scores.

The Ergonomics Analysis Package identified 27 jobs with Concern Level of 4 or 5 (10% of the total). The Rodgers’ tool indicated 16 jobs with High or Very High priorities (6% of the total). The OSHA draft Checklist identified 36% of the jobs as above the threshold for follow-up evaluations. The Rodgers Tool was in agreement with OSHA in identifying 36% of the jobs with a Priority score of 3, 4, and 5. In performing the cross-comparisons among the three techniques, the tools tended to agree on either jobs that were deemed to be a low impact on employees performing the job or jobs that should be studies further; but not both. A sound ergonomics program requires job analysis, medical surveillance and employee input. It appeared that any of the job analysis tools was useful in this context, knowing that surveillance and employee feedback will identify jobs that were missed by the analysis and further study will eliminate jobs that were identified, but not a real concern.

Introduction

Ergonomics is a field that considers the interaction between the worker and his/her job demands. The focus of ergonomics is primarily to reduce unnecessary stress in the workplace. It is this stress placed on the operator by the work demands that can give rise to avoidable injuries and illnesses in the workplace. Because of changes in technology and processes over the past 15 years, workers have been exposed to increased repetitive motion and other risk factors, leading to significant increase in the reporting of cumulative trauma disorders.

The Occupational Safety and Health Administration (OSHA) responded to this "epidemic" with the 1990 Ergonomic Program Management Guidelines for Meatpacking Plants (OSHA, 1990). This set of guidelines has been proposed as a model for "Ergonomic Safety and Health Management" (Hansen, 1993). OSHA divides its guidelines into three distinct parts: (1) management commitment, employee involvement and the development of a written program; (2) four major program elements for a successful plan of action; and (3) detailed guidance and examples for the execution of the program elements. The four major program elements that OSHA embraces are worksite analysis, hazard prevention and control, medical management, and training and education. Worksite analysis is an especially crucial issue since a fundamental aspect of ergonomics is the assessment of jobs with respect to those demands placed on the worker by the job. A workplace survey can therefore aid the employer in reducing occupational injuries and illnesses by identifying existing hazards and conditions that may contribute to such injuries. In its suggestions on carrying out a worksite analysis, OSHA recommends that companies use a "systematic method" or "ergonomic checklist" that includes an analysis of components such as posture, vibration, force and various other cumulative trauma factors. Over the years, many researchers have sought to develop varied ergonomic analyses tailored to accommodate different needs in the workplace. Some ergonomics analyses concentrate on biomechanics and strength limitations of workers. Others consider psychophysical factors in which fatigue plays a role. Still others focus on aspects such as repetition, or a combination of other various factors.

Three Ergonomics Analysis Tools

Three methods of evaluating short cycle tasks were considered for application in an ergonomics program for motor vehicle manufacturing. Two were widely used and a third was developed for the purpose. The two known methods are described briefly here, followed by a more detailed description of the third.

Rodgers Tool

Rodgers proposed a model that considers the physiological relationships between effort levels and contraction and recovery times (Rodgers, 1988). The Rodgers Tool, as it is called here, focuses on body locations prone to overexertion in manufacturing and office jobs. This includes the neck, back, shoulders, arms and elbows, hands and wrists, and legs and feet. It characterizes the stress placed on muscles and joints in terms of force, duration, and repetition. The input data are quasi-quantitative, a judgment based on some subjective decisions and broad quantitative limits. The force was looked at as effort intensity that can be rated as heavy, moderate, or light. This effort intensity was directly related to and categorized by the posture of the body parts in relation to the task undertaken. The duration was defined as the time of continuous muscle effort and was broken up into categories of < 6 seconds of effort, 6 to 20 seconds, and > 20 seconds of effort. The repetition frequency was based on the number of cycles per minute: < 1 cycle per minute, 1 to 5 cycles per minute, and > 5 cycles per minute.

Each of these three factors were then rated for each set of muscles within or across job tasks and from this Rodgers developed a 3-number priority-for-change system, called a Priority Score in this paper. According to the system, different combinations of scores were grouped to determine a given job’s priority-for-change; whether it be moderate, high, or very high. Any combination not shown on the list was a low priority. This system helped to identify the most limiting aspects of jobs in relation to the muscle groups targeted during a given task, and was proposed as a starting point for job improvement discussions. If an analysis of a job results in a high Priority Score, more emphasis was given to the job.

OSHA Draft Checklist

In 1995, OSHA issued pre-proposal materials for the development of an Ergonomic Protection Standard (OSHA, 1995). Although the draft was not officially a Notice of Proposed Rulemaking and was not being promulgated as a rule or standard, the application of the standard would pertain to workplaces where risk factors or work-related musculoskeletal disorders are present. The purpose of the standard was to prevent the occurrence of work-related musculoskeletal disorders by controlling employee exposure to risk factors. The discovery and control of workplace hazards was facilitated by three checklists designed to score risk factors for upper extremities, back and lower extremities, and manual handling tasks.

For the upper extremities (Checklist A) and for the back and lower extremities (Checklist B), job risk factors were listed. The proportion of time during the shift that the worker spends engaging in each risk factor is assessed and a score corresponding to that level of time spent was assigned. Scores for activities averaging 4 to 8 hours per shift were one to two points more than those activities lasting 2 to 4 hours per shift. Any risk factor that a worker performed for under two hours per eight hour shift was negligible.

The materials handling aspect of the job, which has its own (Checklist C), was also a part of the survey. In this section, the relation of the hands to the body during the lift was taken into account. The weight of the load then factored into the determination of a number of points relative to that lift. Other risk factors such as the frequency, one-handed lifts, above shoulder lifts, distance of carry, and torso position during lift were looked at and factored into the total number of points for this section. The score from Checklist C was added to the score for Checklist B. If either Checklist A or the combined Checklist B plus C had a score higher than 5, the job was identified as requiring attention.

Ergonomics Evaluation Package

The Ergonomics Evaluation Package was designed to examine some of the demands placed on workers in automotive manufacturing plants, which are characterized by cyclic work. The result of the analysis was a level of concern that a given job may pose from the point of view of ergonomic design.

The analysis looked at critical body regions separately and compared the demands of the job to the capabilities of workers. The body regions were the back, right and left shoulders, right and left elbows, and the right and left hands. Within each body region, three sets of questions were asked.

• In an instant, can I do this work and will it hurt me? The issues deal with biomechanics and strength.
• In a minute, hour or day, can I sustain the required effort and do I have enough time to recover before the next effort? The issues center around fatigue.
• Over weeks, months or years, will this job cause injury? These issues are cumulative trauma concerns.

Concern Levels were assigned in the following manner. A Concern Level of 1 represented an element or job that can be performed by a vast majority of the working population. It represented a job with no concerns for injury and the job was acceptable. A Concern Level of 2 was an element or job that can be done by most workers with a negligible concern for injury and the job was acceptable. A Concern Level of 3 was assigned to a job or task element that had a potential concern for injury and further study was recommended. Concern Level 4 had a probable level of concern for injury and intervention was recommended. A Concern Level of 5 represented tasks or elements where a concern for injury existed and intervention was required.

To make an assignment of Concern Level, data about the job must be collected and processed. These steps are described here.

On-Site Data Collection

A data collection form was available to record information on the job site. It included spaces to note the job name and location, date, etc. The next section was the place where critical information about the job and job elements were recorded.

To provide a point of reference during the data collection and analysis steps a Sequence Number (Seq #) was assigned followed by a brief description of the task element. To include a task element on the data collection form, one or more of the following characterized the task element.

• Lift or momentary force application of at least 8 lbs
• Sustained hold of an object or force of at least 2 lbs
• Lifting / lowering a part or tool that weighs 8 lbs and moves ³ 10 in
• Repetitive motion
• Contact stress to a body part

Three coordinates were used to record the location of the hands in space. The workspace around the person was referenced to the point between the ankles and was divided into zones. One was the Vertical Zone, in which H was approximately above the shoulders, M between the hips and shoulders, L between the knees and hips and V between the floor and the knees. Another zone was the Horizontal Zone, in which C was the zone close to the body, M the zone that was relatively close and did not require bending to reach, and F the far zone which required an extended reach. The remaining zone was the Lateral Zone, in which F was for the front zone, S for the 90¡ zone on either side of the body, and R for a location that is generally behind the operator. A zone for the task element was based on one of the following four rules.

• If the action was a lift or force application of at least 8 lbs, record the zone as the location at which it is started.
• If the action was a sustained carry or force application including tool use, enter the zone in which most of the time was spent. In this case, the amount of time was also noted.
• If the person was sitting, kneeling or in some posture that appears to change the frame of reference, then the selected zone was that that appeared to be equivalent to one that would represent a standing person with their hips in the same location.
• If the action was a repeated movement, sustained posture or mechanical stress, the notation of zone was not important, but the concern was noted.

Hand, Grip and Wrist Deviation were then recorded for tasks that involve lifting, carrying or sustained force applications. The Hand designators were for right, left or both hands. Grip designators were for pinch, grasp and contact. Pinch was selected for a force application where the finger tips provided the force, which was usually opposed by some other part of the hand such as the thumb or base of the palm. Grasp was selected when the fingers hooked around the object and the force was distributed rather than near the finger tips. Contact was when a force was applied to the object without any further control. Wrist Deviation was indicated as No or Yes for a notable deviation from the neutral position.

Load was the weight of the object, part or tool or the force that was required to perform the task. For instance, if a part was lifted, the weight of the part was entered. If a force was applied, the magnitude of the force was entered. If two hands were involved and the loads were different, the loads attributed to each hand were noted.

For each task, the amount of time that the task element required was recorded. If the time requirement was relatively short, a value of 1 sec was entered. An example was the start of a lift.

The amount of vertical travel involved in a lifting element was noted. If the load was greater than 8 pounds and the vertical motion was more than 10 inches, the distance was entered. A controlled lower was treated as a lift with the starting point based on the position that was farthest from the body in a horizontal direction.

The number of times that the task element was repeated in one cycle of work was then noted. Next was a series of entries for repetitive motions. The first two were used to record repeated motions of the fingers, hands and arms with no impact. No Impact meant that there was little force associated with the movements (e.g., hand starting a nut or screw) or the force application was steady through a distance of more than 1 inch (e.g., using a ratchet wrench). If there was an application of force (e.g., pushing down on a seal or inserting a plug), the number of motions were counted under Forceful, and if a tool was used that had a visible ÒkickÓ to it, the total number of impacts were noted.

For Sustained Posture, when a job required unsupported bending at the waist in excess of 30¡, sustained extension of the neck and/or elevation of one or both hands above the shoulders, the total time during a work cycle for unsupported bending was noted and recorded. In a similar fashion, the total amount of time that the neck was extended to the back or side was recorded. The same was done for unsupported positioning of each hand above the shoulder.

Data Analysis

A spreadsheet was designed to support the Ergonomics Evaluation Package. The first sheet in the analysis was a summary. The remaining sheets were the four analysis sections: Biomechanical, Fatigue, Other Stress, and Lifting.

Biomechanical Analysis Section. The purpose of this section was to look for excessive strength requirements or the potential for excessive forces on the back. The back limits were based on the University of Michigan 2D Biomechanical Analysis with fixed discounts for side and rear zones. The concern levels were based on the percent of a 50/50 mix of men and women that meet the strength requirements for the job element, and in some cases on the compressive limit on L5/S1 disk. The shoulder and elbow limits were based on equations reported by Chaffin and Andersson (1984). The hand grip strengths were based on data reported by Kodak (1983). All of the values were confirmed with other reports and recommendations in the literature.

Considering the most likely task elements to have the greatest demands was sufficient. To do this, emphasis was placed on those elements that required lifting or force applications farthest from the back and those that have the greatest loads. The necessary information was Zone Code (combination of the three letters that described the spatial zone vertically, horizontally and laterally), and the weight or force as distributed between the two hands. There was a separate section for Grip Strength. It considered the grip type (P, G, C), wrist deviation (N,Y), and the weight or force associated with the job element.

The spreadsheet selected the highest Concern Level among the task elements entered for the back, each shoulder, each elbow and each hand.

Fatigue Analysis Section. The analysis looked to see if the force could be sustained for the required time (single effort) and whether there was sufficient recovery in the cycle for all of the efforts required (repeated efforts). The following equation was used to predict endurance time from the effort as the ratio of contraction force to maximal voluntary contraction based on data from Minter (1972).

t_end (min) = 10 ^ (1.0 - 1.96 F_c / F_max)

Required recovery time used the time-weighted average of the relative contraction force with respect to the median strength of a 50/50 mix of workers. RohmertÕs equation for recovery time was adapted to determine a total cycle time that was needed to assure adequate recovery (Rohmert, 1973).

t_cycle = t_c + 18 (t_c / t_end)^1.4 (F_c / F_max - 0.15)^0.5 t_c

Using these two equations meant that for any given value of F_max, t_cycle could be determined for t_c. In other words, the equations answered the question: How much recovery time should be allowed after a muscle contraction? In the analysis of work that has repeated job elements, the question that must be asked is: Is there enough recovery time within a cycle of work to prevent muscle fatigue?

To illustrate how the Ergonomics Evaluation Package Fatigue Analysis approached these questions, consider a generic muscle group that has the strength (maximum force of contraction, F_max) of the average person. During a cycle of work, it may be loaded (under contraction) one or more times. In general, the greater the load time the less the force of contraction can be. Alternatively, as t_c increases the overall cycle time (t_cycle) must increase to allow sufficient recovery. Used another way, for a given t_c and t_cycle, the greatest force that the muscle can exert and still have sufficient recovery time can be determined. The first step was to determine the total amount of time under load. The total load time and the cycle time were then used to determine a maximum loading, which was the percent of F_max that can be sustained for that period.

F_max was selected to be the strength of the average worker for each muscle group associated with each body part in each posture in the Ergonomics Evaluation Package. The next step was to compute the loading for each exertion, which was the percent of the actual force to F_max for that body part and posture. Then the time-weighted average (TWA) of the loadings for each body part was computed. This TWA-Loading was the time-weighted average of the percent of F_max. If the TWA-Loading was less than the maximum loading, then the respective muscle group for an average person should not fatigue

Because the strength (F_max) for individuals will range above and below that of the average, the Concern Levels were selected such that Level 1 assumed a strength of 71% of the average, which was the approximate strength of the 5th %ile of a 50/50 mix of men and women. Level 2 was bound by 85% of the average, which was the approximate strength of the 25th %ile. Levels 3 and 4 were bound by 120% and 168% of the average, which were the approximate strengths of the 75th %ile and 95th %iles respectively.

For the actual analysis, task elements that were important were those in which objects or forces greater than 2 lbs were exerted or an unsupported posture was sustained. Again the Sequence Number was noted followed by the Zone Code, the distribution of load between hands, the Grip Code, wrist deviation (N,Y), and the time that that task element was sustained in one cycle.

There were two special cases. One was the unsupported bend and the other was hands above the shoulders. The total time in that posture was used in the analysis. The spreadsheet computed the Concern Levels for the back, each shoulder, each elbow and each hand. It placed the results on the Summary Sheet.

Other Stress Analysis Section. The next analysis sheet was for stresses associated with repetitive motions and neck extensions. One area was the number of repetitions that the job required of the hands and arms. For Right and Left, the number of repetitions associated with each hand/arm was considered. If there were tool kicks or other impacts, this was factored in. This section also assessed the degree of neck extension risk by the percent of cycle time. The Concern Level associated with repetitive motions and neck extension were computed and passed to the summary sheet.

NIOSH Lifting Analysis Section. If the NIOSH lifting analysis was to be performed, this section was selected.

The lifting analysis data entry is similar to the previous work sheets. The Concern Level was computed as a function of the Lifting Index following the NIOSH method for multiple lifting situations and entered on the Summary Sheet.

Comparison of Methods

Data were collected during the months of December 1995 and January 1996 on 267 jobs in the final assembly area of an automotive plant. Four hours of training were provided in a classroom setting prior to the field evaluation. On arrival at the site, data sheets were filled out as a team in order to ensure consistency within the group. After this initial on-site training, individual team members were responsible for gathering data for a job. The team members observed the worker at the job and completed the worksheets for the Ergonomics Analysis Package first, and then the Rodgers Tool and OSHA Checklist. If a particular job was identical for the right and left sides of the assembly line, only one set of data sheets were completed and the sheets were so designated as referring to both the right and left sides. Loads for tools and other objects lifted were measured with a mechanical force transducer. Stopwatches were used to time the workers during sustained postures and during lifting and carrying tasks. In addition, jobs were videotaped to include at least two cycles of a worker performing a task.

The team members separated into groups of two in order to form consensus data sheets for the Ergonomics Analysis Package, the Rodgers Tool data sheet and the draft OSHA Checklist. To help facilitate the formation of the consensus sheets, the videotape for the job was reviewed along with the process engineering stack sheets. The data for the Ergonomics Analysis Package consensus sheets were entered into a spreadsheet written in Excel, version 5. The results were summarized on the first page of the spreadsheet by body location and type of stress. For each combination of location and type of stress, a Concern Level with values for 1 to 5 was assigned. The overall score for the job was the highest individual Concern Level.

The original Rodgers Tool data sheet for job analysis contains a section termed "priority for change". The various combinations of scores stem from the three categories of force, duration, and frequency of a task at various levels of intensity. For our purposes, a priority score of 3 was given to tasks in the moderate category. A priority score of 4 and 5 were assigned to tasks deemed at a high or very high priority for change, respectively. For those jobs which had no priority category for the Rodgers’ Tool, a score of 1 was given.

The revised draft OSHA Checklist was divided into three sections for analysis. These sections focused on risk factors for cumulative trauma, back stress, and manual materials handling. A total score for each section was computed based on the values OSHA assigned to the different risk factors. The scores for Checklist B and Checklist C were combined to give a total score for Checklist B/C. If either Checklist A or Checklist B/C had a score higher than 5, the job was considered to be a ÒflaggedÓ job.

The amount of effort to use the three methods is mostly reflected in the time it takes to complete them. The on-the-floor time for gathering data for the Ergonomics Analysis Package was approximately 25 minutes. Another 20 minutes was spent developing consensus data sheets for the Ergonomics Evaluation Package and the time needed to complete the spreadsheet portion of the Ergonomics Evaluation Package was approximately 10 minutes. The total time, therefore, spent for the Ergonomics Evaluation Package was about 55 minutes for each job. On the other hand, the completion of the on-site evaluation was 10 minutes each for the OSHA Checklist and Rodgers Tool. Developing consensus data sheets for these two tools took another 5 minutes each, giving a total of 15 minutes spent for evaluating jobs using the OSHA Checklist and the Rodgers Tool. This is about one-quarter of the time that it took to analyze the jobs with the Ergonomics Evaluation Package. In order to justify this additional time, the Ergonomics Evaluation Package should be able to provide more information, accuracy or both.

Individual Analyses / Staged Outcomes

The distribution of outcomes for the Ergonomics Evaluation Package and Rodgers Tool were examined. Looking for the worst jobs (those with a Concern Level or Priority Score of 4 or 5), the Ergonomics Analysis Package and Rodgers Tool were similar at recommending 10% and 6% of the jobs for immediate attention, respectively. Although the values were relatively low, the higher number of ÒredÓ jobs identified by the Ergonomics Evaluation Package would suggest that it was more protective than the Rodgers Tool for red jobs.

Using a lower threshold for an alert (yellow jobs), an Ergonomics Analysis Package Concern Level of 3 and a Rodgers Tool outcome at Moderate Priority, the two schemes pointed out 36% and 30% of the jobs, respectively. Again, the Ergonomics Evaluation Package appears to be more protective. As a combined threshold (yellow/red), the Ergonomics Evaluation Package resulted in 46% of the jobs requiring a closer examination, while the Rodgers Tool resulted in 36%. That is, the Ergonomics Evaluation Package tended to flag about 10% more jobs than the Rodgers Tool did.

There was no gradation of the OSHA Checklist for yellow or red jobs, but the OSHA Checklist did agree with the Rodgers Tool in suggesting that 36% of the jobs required a closer examination (yellow/red), or that 64% of the jobs were acceptable. The Ergonomics Evaluation Package, however, was somewhat lower with 54% of the jobs being acceptable because it leaned toward the protective side for yellow/red jobs.

Inter-Comparisons

There was no standard or independent measure of job demand to compare the outcomes to for this study. Inter-comparisons among the three techniques were therefore looked at.

Overall Outcomes: Yellow/Red versus Green

The division of jobs into green and yellow/red categories was useful to understand what jobs did not require any further attention and those that may but with no further prioritization. The inter-comparison of the three analysis schemes for yellow/red jobs showed some consistency between the OSHA Checklist and Rodgers Tool at about 56% of the time (Sensitivity = 0.56). For jobs considered to be low impact by the OSHA Checklist and the Rodgers Tool, they saw about three-fourths of the jobs as acceptable (Specificity = 0.75). That is, they agreed much more on what did not require further attention.

When the Ergonomics Analysis Package was compared to either the OSHA Checklist or Rodgers Tool as the standard, the Ergonomics Evaluation Package saw about 70% of the jobs as being in the yellow/red category when the other tools did (Sensitivity Å 0.7). In addition, approximately 70% of the jobs were considered by the Ergonomics Evaluation Package to be acceptable when the Rodgers Tool or OSHA Checklist considered them to be acceptable (Specificity Å 0.7). And with the Ergonomics Evaluation Package as the standard, the OSHA Checklist and Rodgers Tool ÒcorrectlyÓ saw 54% of the jobs as yellow/red (Sensitivity = 0.54) and about 80% of the jobs as not yellow/red (green) (Sensitivity = 0.8).

No matter what method was taken as the standard, there was good agreement among the three methods about what jobs were green. Interestingly, the Ergonomics Evaluation Package as the trial method agreed fairly well with the Rodgers Tool and the OSHA Checklist with regard to yellow/red jobs while OSHA and Rodgers did less well between themselves or as the trial method in comparison to the Ergonomics Evaluation Package as standard.

Overall Outcomes: Red versus Green/Yellow

Another purpose of an ergonomics evaluation was to identify jobs that require immediate attention versus those that do not require any or delayed attention. For the jobs in the red category versus those in the green/yellow, the Ergonomics Analysis Package and Rodgers Tool tended to agree on jobs that were deemed not to be a high impact on employees performing the job. With either one of them as the standard, the agreement on the lower priorities (specificity) was high, in the 90% range. The sensitivities, however, were low with either one of them as the standard. That is, they did not agree well on what jobs required immediate attention. Because there was no standard, it is difficult to conclude whether either one is better at selecting high priority jobs.

Conclusions

Three methods (Ergonomic Analysis Package, Rodgers Tool and OSHA draft Checklist) for assessing the work demands on employees were used on 267 jobs in a final assembly area of a automotive plant. All three analysis tools agreed that about two-thirds of the jobs studied did not require further attention. The three tools did not agree, however, on the jobs that would require any type of further attention, either as low or high priority.

Practically speaking, the OSHA Checklist and Rodgers Tool were similar in that they: (1) needed the same amount of time to complete (about 15 minutes), (2) both saw the same percentage of jobs as okay and having some need for further examination, and (3) compared equally against the Ergonomics Analysis Package. This result was not expected since they were much different in their approaches. Caution must be used with this conclusion, since there was no way to determine the percentage of high priority jobs for the OSHA Checklist. It was possible that although the two tools saw the same number of jobs as acceptable, they might differ sharply on what jobs need immediate attention. While the Ergonomics Analysis Package required more time, its principle advantage was the ability to clearly focus on the major contributors to the outcome.

A sound ergonomics program requires job analysis, medical surveillance and employee input. It appeared that any of the job analysis tools was useful in this context, knowing that surveillance and employee feedback will identify jobs that were missed by the analysis and further study will eliminate jobs that were identified, but not a real concern.

Acknowledgments

Important support was provided by Ford Motor Company and AutoAlliance International and many employees with each of these organizations. At USF, Michael Alexander, Gregory Deal, Michelle Logan, Perry Logan, Nanda Kittusamy, and Paul McCright come immediately to mind as providing essential help along with the support of many others.

References

Chaffin, DB; Andersson, G (1984): Occupational Biomechanics. New York: John Wiley & Sons.

Hansen, JA (1993): OSHA regulation of ergonomic health. Journal of Occupational Medicine 35:42-46.

Kodak (1986): Ergonomic Design for People at Work II. New York: Van Nostrand Rienhold.

Minter, AL (1972): Comments on ÒAssessment of isometric muscle contractionsÓ. Ergonomics 15:453-455.

Occupational Safety and Health Administration (OSHA) (1990): Ergonomics Program Management Guidelines for Meatpacking Plants. OSHA 3123.

Rodgers, SH (1988): Job evaluation in worker fitness determination. In Himmelstein, JS; Pransky GS eds: Occupational Medicine: State of the Art Reviews.

Rohmert, W (1973): Problems of determination of rest allowances, Part 2: Determining rest allowances in different human tasks. Applied Ergonomics 4:158-162.