"Any validity study should be based upon a review of information about the job for which the selection procedure is to be used. The review should include a job analysis..."

Uniform Guidelines 1978

There is an established legal need [in the United States] for job analysis. In an article entitled Selection's Uniform Guidelines Help, Hindrance, or Irrelevancy? Christopher Daniel points out, "Courts continue to decide selection cases, often deferring to The Uniform Guidelines on Employee Selection Procedures" (1989, p.68). If an employer wishes to demonstrate to the courts that the selection process used for an employment decision was valid, the employer will need to start from the basis of a current job analysis.

Not only are there legal reasons for job analysis but there are also important management considerations as well.

The desire by business and government to manage work and the human resources of an organization requires the study, analysis, and organization of work activities which make up jobs. The need to manage work activities established a need for job analysis before the Uniform Guidelines called attention to its legal significance.

Job Analysis Principles

It is important to take note that the Uniform Guidelines (1978) state, "Any job analysis should focus on the work behavior(s) and the tasks associated with them. If work behavior(s) are not observable, the job analysis should identify and analyze those aspects of the behavior(s) that can be observed and the observed work products" (1978, p.38302). This description is very important for establishing what is meant by job analysis. According to the Guidelines, job analysis is a record of observable behaviors or observable work products. Job analysis is not a record of thought processes, attitudes, traits, constructs or initiatives. These worker traits are outside the restrictions of observable behaviors or observable work products.

A number of fundamental principles concerning jobs and the process of analyzing them have been identified.

These principles are:

1. All jobs can be analyzed and recorded.

2. Job analysis can enhance communication.

3. The process of analyzing jobs can easily accommodate change.

4. The process can be clear enough so employees and employers can understand and contribute to the process.

5. The process can be designed so that all major personnel decisions can be based on the resulting data.

6. Skill, knowledge, and ability can be defined in operational terms.

7. Job analysis based on observable behaviors and work products contributes to efficient human resources management.

8. Nearly everything that needs to be written to explain the work of a job is already written.

1. All jobs can be analyzed

Job analysis is the process of defining the work, activities, tasks, products, services, or processes performed by or produced by an employee or employees. Often employees express the opinion that they know what they do; but it is too complex, or too abstract, to be written down or defined. Based on experience of analyzing managerial, technical, clerical, and labor jobs, this assumption is not accurate for the following reasons. The work of all jobs is performed by people. People do not behave in random fashion. People are reasonable and systematic. People are continuously adjusting their jobs to be more efficient. People, when asked, can talk about their jobs in a very organized, systematic manner. When employees review a written job description of their duties three to six months after the analysis was performed, they readily recognize the description as their job. Finally, all employees at every level in an organization are trained to do their job. That training could not occur if it were not possible to analyze and define jobs.

2. Job analysis can enhance communication

Job analysis should respond to that basic question asked by all employees at one time or another, "what do you want me to do"?

If the job analysis does not improve communication and the understanding of what is to be done, it impedes communication and may be more detrimental than beneficial to the organization. Good quality job analysis can and should improve communication in the organization.

3. The Process Can Accommodate Change

Job analysis has been and still is tedious work. Organizations tend to avoid conducting job analysis because of the amount of resources necessary to do it and do it well. Jobs are dynamic and constantly undergoing change. Job analysis systems have to address this fact of change. Data about jobs that are out of date are not valid and may be misleading. If the data are not valid, any employment decisions based on that data may be suspect. Job analysis data should periodically be reviewed by employers and revised to reflect the changes which occur over time.

The system for analyzing jobs should be designed to take full advantage of computer technology. Job analysis is an ideal use for computers. Changes over time can easily be made to the data base and updated job descriptions produced relatively easily and quickly. Technologically there is no reason for an organization not to have current, comprehensive job analysis data and job descriptions.

4. Employees and Employers Can Understand the Process

Based on the experience of analyzing many different jobs at a variety of levels in an organization, it has been established that employees can identify anywhere from a hundred to three hundred tasks per job.

Employees and employers understand that these tasks will vary widely in terms of level of difficulty and the amount of time required to perform them. When provided a relative scale, employees have been able to rank hundreds of tasks in terms of relative frequency and relative difficulty. Relative frequency and relative difficulty provide an excellent method for defining critical aspects of the job.

With hundreds of tacks, when scored for difficulty and frequency, there will result a great deal of data on a single job. If there are proficiency statements or quality dimensions set for a large number of the tasks, the data will further increase. If minimum skill levels are set for the tasks, there will be additional data generated for each job. If there are 100 jobs in an organization, it is clear that job analysis will generate a tremendous amount of data. It is, therefore, important that each of the steps or processes be reasonable and simple.

5. Job Analysis Can Be the Basis for All Major Personnel Decisions

Quality job analysis data can and should contribute greatly to; selection and hire, performance evaluation, training and development, compensation, job design, work force projections, and work force reduction or expansion decisions. There should be an audit trail from the job analysis to the decision. The selection of an individual should be based on the individual's ability to do the critical tasks of the job.

Performance evaluation should also have an audit trail to the critical tasks of the job. There should be quality standards or level of proficiency statements for critical tasks of the job. The primary question of performance evaluation is, did the employee do what the employee was asked to do? The secondary question is, did the employee do it at an acceptable quality level? If the information is in the job analysis, both what is to be done and at what quality level then these questions should be relatively easy to answer.

Compensation decisions should have an audit trail back to the critical tasks. There are a number of factors which may be used to determine relative compensation. A question that needs to be answered for setting relative compensation in an organization then is: What skills are required to perform the work of various jobs? Jobs requiring higher skill levels usually receive higher levels of compensation. There are other factors used in setting compensation, of course, such as, the external labor market, longevity, and working conditions; but relative skill level is very important for setting relative compensation. The job analysis process should establish relative skill levels among the jobs in an organization.

A job analysis should be an essential element to the development of an efficient training program. It should serve as the basis for the needs analysis which is the first step in developing training programs. It can be used to identify the different skills needed for each job and the marginal skill differences between different jobs. Employees, as well as the employer, should understand the marginal skills necessary to be proficient at different jobs in an organization. Information about the marginal skills will help employees plan their own professional development. It will also help the employer develop training programs to meet skill shortages in various jobs.

Detailed job analysis is essential to good work force planning and utilization. As organizations are forced to react to an ever-changing economy, they will often be confronted by questions of reduction or expansion of their work force. It is important that organizations identify core functions that must be maintained and the skills required to carry out these functions. It will also be necessary for companies to have plans in place on how they are going to staff up rapidly to take advantage of favorable economies and how they will reduce and consolidate jobs when the economy slows down.

6. Skills, Knowledge and Ability Can Be Operationally Defined

Skills, knowledge and ability can and should be operationally defined. Skills, knowledge, and ability have been widely used in a variety of applications in personnel work in the past. These terms are used in determining relative compensation and in the area of human resources development, as well as other areas of human resources management.

7. Job Analysis Based on Observable Behaviors and Work Products Contributes to Efficient Human Resources Management

By recording behavior and work products in the job analysis, an organization concentrates its attention on those aspects which the organization and the employee can manage. They can be understood and measured by both parties. Training programs can be developed based on these tangible elements of the work, and learning can be measured. For example if the job requires an individual to drive a truck, it is possible to develop a truck driver training program and then have the person demonstrate their truck driving ability to measure whether or not they learned how to drive a truck. This approach of concentrating on the tangible elements of the job avoids becoming immersed in the less tangible aspect of jobs such as traits and constructs.

The real saving to any organization is selecting a skilled person the first time. There is a much greater probability of this happening by concentrating on behaviors and work products than by concentrating on traits and constructs unless they have been thoroughly validated.

8. Everything That Needs To Be Written for the Job Is Already Written

Introduction

The data collected from the job analysis has to serve many purposes. It is important to know how the data will ultimately be used since that will affect, not only the data collected, but also how it should be recorded and stored. If the data are to be used only for selection purposes, then the data collected should be sufficient to meet that need. If it is to be used for performance evaluation, training and development, compensation, and job structuring in addition to selection, then the uses for the data will be much greater.

With the advent of the personal computer and the availability of powerful software programs, much of the tedious work of job analysis can now be more efficiently managed. Like the binary principle of the computer, job analysis can be the collection of lots of simple data in a well organized manner rather than the collection of small amounts of very complex data.

Managers as well as those employees who perform the work of the job need to understand the process and the logic upon which the job analysis process is based. They need to have confidence that the resulting data from the process will represent the work of the job. The initial data collection, the interviews of incumbents and supervisors, the designation of critical tasks, the assignment of skills, knowledge, and ability, and quality standards all need to be reasonable and understandable to all parties. This reasonableness is a central concern for job analysis and for good human resources management.

The Interview (Initial Data Collection)

There are two different types of jobs: the newly created job and the job with a history. The two types of jobs call for slightly different methods of job analysis. The newly created job is different from the established job in that it has no history. It requires speculation, albeit informed speculation. There is no incumbent from whom to gather data. The job analysis of the newly created job should be conducted with the person creating the new job or the person who is to supervise the new job. The initial data collection process should be similar to that for an established position.

The initial data collection of the established job is achieved through interviews with one or more incumbents in the job. It is assumed that incumbents are the most knowledgeable people about the duties and responsibilities of the job. It is possible that an incumbent may not be knowledgeable of all aspects of the job or may even be doing the job wrong. There will be opportunities for management to monitor the data at a variety of steps in the process to correct for errors. If there are a large number of people who do similar jobs, it is not necessary for all of them to be involved in the initial interview. Three to five people who are very familiar with the job seem to work well. The initial interview data should be entered into a computer. A computer printout of the initial interview data can be distributed to everyone in the job, so they can revise the data according to their perception and experience with the job.

The job analyst or interviewer asks the incumbents to talk about what they do. Incumbents should be informed that they will have more than one opportunity to provide information on their work. Should they fail to mention something during the first interview, there will be additional opportunities to revise the data.

If the interviewer does not understand what the incumbent is talking about, it is very important to ask for clarification. Terminology is often unique to the job and subtle differences can be significant. It is also important for the interviewer to watch for duplication and overlap. Terminology and processes can creep into a job and take on an existence of their own. If the analyst or interviewer asks clarifying questions, the incumbents should be able to explain why things are done as they are on the job. When an interviewee defends a process with the statement, "that's the way we've always done it," the interviewer should make a note to check out the rationale for the process or procedure with another person such as a supervisor. Incumbents should be provided plenty of time to talk about their job. In some cases, this initial interview may take up to four hours. If there is more than one incumbent in the initial interview, the incumbents should be encouraged to talk among themselves during the interview. The primary concern is to record as accurately as possible the tasks the employees perform.

Management should review the data from the initial interview to answer the questions:

1. Does the data reflect the work being done by the employees, and
2. Are the employees doing what they should be doing?

Managers and supervisors who are responsible for the quantity and quality of the work need to know that they are an integral part of the job analysis process.

Verify The Information

Once the initial list of tasks has been compiled, the list should be sent to all of the incumbents in the job. It provides an opportunity for all employees to have input into the process. Frequently, additional information will come to light about the job. Jobs are dynamic and ever changing. Different patterns of how to do the work and different activities may be discovered. At this point in the process, it is better to err in the direction of including more information rather than less.

The Uniform Guidelines for Employee Selection uses the phrase "work outcome(s) and measures of their criticality and/or importance" (1978, p.38304). There can be a number of variables used to establish this "criticality." Some of the more common are: time, frequency, difficulty, and consequence of error.

Relative frequency and relative difficulty are recommended. Once the list of tasks has been agreed to, the next step is to ask the incumbents in the position to score each of the tasks for relative frequency and relative difficulty. The incumbents are given a copy of the scoring instructions and told to score each task for frequency and difficulty with a number from 0 to 7. Since these are relative scores, the employee needs to identify those tasks which are performed most often relative to all of the tasks in the job and give them a score of 6 or 7. Next, the employee is asked to identify those tasks which are performed infrequently relative to all of the tasks in the job, and score them 1 or 2. Since these are relative scores there should be approximately the same number of tasks with high scores as tasks with low scores.

Once the employees have scored the list of tasks for frequency and difficulty, it is a simple matter to establish which tasks are critical. The employer uses these scores as a guide to establish a cut-off score. A cut-off score of four or five is simply a statement by the employer that, if the employee can perform the critical tasks, those with combined scores of five or six and higher; then it is assumed the employee will be able to perform the lesser tasks.

Once the scores are received from the incumbents in the job, they can easily be entered into the computer. The computer can sum frequency and difficulty for each task. It can sum and average the frequency and difficulty columns and the combined column. The computer can then sort those tasks which meet or surpass the cut-off score and print out the list of critical tasks. In some cases management may elect to designate high scoring tasks as non-critical tasks because only a small number of employees perform these specialized tasks. It will be the list of critical tasks which will be the basis for all future personnel decisions.

Skills, Knowledge and Ability

As stated in the Uniform Guidelines, "In the case of a selection procedure measuring a knowledge, skill, or ability, the knowledge, skill, or ability being measured should be operationally defined" (1978, p. 38302).

Determining relative compensation for jobs within an organization is without a doubt a very important function. In order to make this relative compensation determination among jobs, a valid process will be necessary to measure each job with the same criteria. It will also be necessary to establish what it is in each job that justifies the rating it receives.

The Skill Taxonomy

Each of the jobs in the organization should be measured by the same criteria when establishing minimum skill requirements. An organization can develop a list of skills, a skill taxonomy.

Recruitment and hire

This model of job analysis clearly establishes the skills needed for a job. It also establishes the different skills required for each of the jobs in the organization. For example if a company has a series of jobs such as Building Maintenance Worker I, II, and III, this process of job analysis will not only establish the skills required for each of the jobs; but it will also establish the different skill levels, "marginal skill levels", between each of these jobs. It is then possible to develop training programs based on the specific tasks unique to the job as well as those similar tasks required by other jobs.

At a time when jobs are changing relatively quickly, this information about marginal differences between jobs will be invaluable to the development of efficient, effective training programs. This information about marginal skill requirements and marginal behaviors and work products will be helpful in the maintenance of high-quality human capital for an organization.

Compensation

Another benefit from this approach to job analysis is that it can signal when compensation for a job should be reviewed. Jobs are dynamic and always changing. The question of when these changes are significant enough to result in a change in compensation is important. Since this job analysis model is designed to take advantage of computer technology, it is relatively easy to periodically update the tasks of jobs. Tasks of a job may change but that does not necessarily mean that the minimum skills have changed. Relative frequency and relative difficulty may change but that does not necessarily mean that minimum skills have changed. When the change in the job results in a change in critical tasks which result in changes to minimum skill levels, then the job should be reviewed to determine if compensation for the job should be adjusted. This process provides documentation to support adjustments to compensation based on changes in minimum skill levels.

Work force planning and utilization

This approach to job analysis will identify which tasks in a job require the highest skills. Management will be able to use that information to structure jobs. If there are only a few tasks requiring high skill levels, it may be more efficient for the organization to shift those tasks to another job which already requires those skills. This allows the organization to develop specialized jobs and hire employees with an emphasis on those skills. The corollary is the concept of hiring generalists. The company may want to develop employees who are competent in a variety of areas and will be able to perform a wide variety of tasks. In many organizations, there will be a need for both specialists and generalists. This model approach to job analysis will help the organization document where the specialists are and why, as well as where the generalists are and why.

The organization will be able to demonstrate job ladders and lattices which is very important for career planning by employees as well as the employer. Employers will be in a better position to show which jobs require on-the-job training, which require employer sponsored training and which require training through some other institution. Employees will have access to very important information to decide whether or not they want to expend the time and energy needed to obtain the skills for other jobs in the organization.

The organization can also use the job analysis information for projecting work force needs. By observing how jobs change over time, an organization will be in a better position to predict how jobs are likely to change in the future. Employers should be in a better position to project lead time needed to bring an employee to the level of full productivity. As organizations decide their future human resource needs, the information obtained through this job analysis model should aid the organization in developing a plan to meet those needs.

Summary

The Uniform Guidelines attached a significant legal importance to the process of job analysis. The Guidelines expanded the role of job analysis to all major employment decisions. Job analysis is necessary for selection, training, compensation, performance appraisal and separation decisions.

A job analysis process needs to be designed to meet all of these needs. In addition to being legal, a job analysis process should improve communication, should accommodate change, should contribute to improved human resource management and should be cost effective.

The process needs to improve communication. It has to be understood by employers and employees. It needs to be a tool for both parties in managing their work and their future. The process needs to be valid, it needs to do what it says it is doing. Employees and employers need to know they are using the same definitions and the same measures. Concentration should be directed at observable behaviors and work outputs and away from abstract concepts which are difficult to define and measure.

Jobs are dynamic and undergo constant change. A job analysis process needs to be designed to accommodate that change. Job analysis is an ideal application for the computer. By recording the data in a computer, adjustments in the job can be easily recorded and updated. Changes in the work of the job can be examined to determine if the change justifies adjustment in job structuring, training or compensation. These changes then can be integrated into the human resources management system when they occur rather than waiting an extended period of time to do an organization-wide job analysis. Gradual change is much easier for an organization to absorb than massive change every five or ten years. It is important for the job analysis model be planned to accommodate change.

Job analysis should assist an organization in obtaining and retaining high-quality human resources. It should help the organization manage its human resources development program. It should help the organization define career ladders and lattices. Both the organization and the employees should be in a better position to manage careers as a result of the job analysis process.

References

Christal, Raymond E. The United States Air Force Occupational Research Project, Occupation Research Division, Air Force Human Resources Laboratory (AFSC) Lackland AFB, Texas, 1974.

Daniel, Christopher, Selection's Uniform Guidelines Help, Hindrance, or Irrelevancy?, Review of Public Personnel Administration, Spring 1989.

Gail, Sidney, Job Analysis, San Fransico, 1987.

Gilpatrick, Eleanor, A lob Analysis Method for Developing Job Ladders and for Manpower Planning, The Research Foundation, City University of New York, 1971.

Kearney, Thomas I., Bessie J. Kuhn, Irene Lopez, Earl Lee, William Machintosh, Harry Nussberger, and Barbara Taylor, Handbook For Analyzing Jobs, Washington D.C. 1972.

Manese, Wilfredo R., Occupational Job Evaluation A Research-Based Approach To Job Classification, New York, 1988.

National Task Bank, Tasks in Social Welfare, U.S. Department of Health, Education, and Welfare Social and Rehabilitation Service State Manpower Development Staff, Washington D.C. November 1975.

Uniform Guidelines on Employee Selection Procedures (1978), Federal Register, Friday August 25, 1978 Part IV.

Yoder, Dale, Herbert G. Heneman, Jr., Staffing Policies and Strategies, ASPA Handbook of Personnel and Industrial Relations, Vol. 1, Washington D.C., 1974.

Appendix A List of Tasks with Frequency, difficulty and Skill Scores

TASK LIST-With Frequency, Difficulty and Summary Ratings-With Maximum Skill Level Ratings

By James P. Clifford James P. Clifford is a Partner in the Human Resources Systems Institute. He has worked in city government for over twenty years, holding the positions of Human Resources Director, Grants Management Director, and the Director of Employee Development. He has a Master' s Degree from Michigan State University in Labor and Industrial Relations and teaches a graduate course in Human Resources Management. He is an active member of International Personnel Management Association, American Society of Training Developers, American Society for Public Administration, and the Industrial Relations Research Association.