Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
6 Procedures Used to Produce the Fourth Edition Dictionary of Occupational Titles Because the Dictionary of Occupational Titles is a dictionary and a reference manual, and because it is widely used and relied on to provide comprehensive, authoritative information about occupations, it is especial- ly important that the information it contains be reliable and accurate. The comprehensiveness, reliability, and accuracy of the DOT are in large part a function of the data collection and analysis procedures used to produce it. This chapter describes these procedures step by step, from the initial selection of industries for study to the final stage of writing definitions for the occupational titles that appear in the published volume. This description, along with the more technical evaluation in chapter 7 of the implications of these procedures for the quality and characteristics of data contained in the DOT, is intended to provide information that will enable users of the DOT to make informed judgments about its value. In the course of our analysis it became clear that the production of the DOT is seriously underdocumented. Because of the lack of published technical information or documentation of procedures, the description in this chapter draws heavily on information gathered during interviews with staff at the national office and at seven field centers (Arizona, California, Michigan, Missouri, North Carolina, New York, and Texas) as well as on Pamela S. Cain had primary responsibility for the preparation of this chapter. ~4
Procedures Used to Produce the Fourth Edition DOT ~5 information found in instructional manuals developed primarily for internal use, the Handbook for Analyzing Jobs (U.S. Department of Labor, 1972) and the Definition Writer's Manual (U.S. Department of Labor, 1974~. HISTORICAL BACKGROUND The procedures used to produce the fourth edition DOT were originally developed by two different research programs dating from 1934 and 1949. The first program of occupational research was initiated in response to the depression. At that time the large number of unemployed, qualified workers made job matching on the basis of work performed, by comparing an applicant's work history with a job order, relatively easy. Occupational research at this early stage hence focused primarily on job content. The research begun in 1934 formulated major principles of job analysis methodology for the DOT and set in place sampling techniques and on-site procedures for the study of business establishments that are still used in the program today. World War II and postwar economic recovery prompted an influx into the labor force of workers who had little or no prior labor force experience or up-to-date job skills. Increasingly, it was necessary to match jobs to workers on the basis of an applicant's potential to perform a given job rather than his or her demonstrated expertise. In 1949 the functional occupational research project was begun to investigate systematically various dimensions of worker potential. This research, which culminated in the third edition DOT (1965), introduced worker functions in relation to data, people, and things into the DOT classification scheme, standardized definition writing, and developed various scales for rating worker traits. Since publication of the third edition DOT in 1965, few, if any, modifications appear to have been made in the basic methodology of the occupational analysis program. SAMPLING FOR THE DOT Sampling for the fourth edition DOT was by no means straightforward. The sampling universe was all jobs in the national economy. Conceptually, the economy was categorized by occupations. For purposes of collecting data about occupations, however, the economy was categorized by industries; that is, industries were the basic units by which assignments were made to the field centers. Once field centers received their industry assignments, they were responsible for obtaining complete coverage of the jobs within their assigned industries. In order to study jobs, however, the business
116 WORK, JOBS, AND OCCUPATIONS establishments in which jobs are found had to be located. Only after this had been done were jobs finally selected and analyzed. This multistage selection process identified an increasingly disaggregat- ed unit of analysis; the process then shifted direction: for purposes of compiling the DOT, data about jobs were aggregated to form occupations. Occupations in the DOT are not intended to be either firm or region specific, and in some cases not even industry specific. Rather, the descriptions are intended to reflect the occupation as it typically occurs in the national economy, although what is meant by "typical" is not explicitly defined. The various steps involved in the creation of the fourth edition DOT, from industry assignment to occupational composite, are described in the subsequent sections of this chapter. ASSIGNMENT OF INDUSTRIES TO FIELD CENTERS Shortly after the completion of the third edition DOT, industry assignments were made for the fourth edition. Most centers were assigned the same set of industries for which they were responsible in the third edition. Although there is no documentation of how assignments were made, in matching centers and industries, national office staff appear to have relied on a combination of common sense and general knowledge as well as on their own experience in occupational analysis. It was impossible to determine conclusively the basis or criteria of industry assignment. When national office staff were queried as to how assignments had been made, responses were vague. North Carolina, for example, was said to be the "logical place for machine textiles and wood furniture"; similarly, Washington was said to be "the natural place" for aeronautics, given the location of Boeing in Seattle; while it was "pretty obvious that Texas doesn't have logging." For relatively widespread or highly visible industries the rationale for assignment was of this common sense variety. For other, more obscure, industries the assignment of which was not so obvious, various documents were consulted. The annually published Thomas Industrial Register, for example, was consulted frequently for this purpose. It was said to be of limited usefulness, however, because it lists a company's home office rather than its production sites. In addition, various state and local industrial or manufacturing directories were consulted on a case-by-case basis, as were assorted federally produced censuses of manufacturers and businesses. Generally, industries were assigned to centers in whose state or region they were thought likely, or sometimes actually determined, to be located. Large industries, such as banking and financial institutions, were usually assigned to more than one field center to ensure adequate coverage and to
Procedures Used to Produce the Fourth Edition DOT ~7 detect regional variations in jobs. Groups of related occupations that are not industry specific (e.g., the "any industry" designation) were assigned en bloc to a given center. Assignments were made using the 229 industry designations of the third edition. The number of industries assigned per field center is given in Table 6-1. Fourteen industries (6 percent of the total) were unassigned, and no new job analyses were collected from them; 40 percent were assigned to one field center; and the remaining 54 percent were assigned to two or more field centers. On average, each center was responsible for 42 industries, ranging from 11 for the Arizona special project to 84 for California. Certain field centers appear to have had greater responsibility for industry coverage than others. California, for example, was assigned the largest number of industries and was solely responsible for 9 percent of the 232 industries in the fourth edition DOT industry designation. New York had the second largest number of industries and was solely responsible for 8 percent of the total. Washington was assigned a smaller number of industries and had sole responsibility for TABLE 6-1 Field Center Industry Assignments Number of Number Industries Uniquely Center a Assigned Assigned Arizona (2) 11 1 California (16) 84 20 Florida (8) 28 0 Michigan (12) 35 17 Missouri (12) 62 13 New York (18) 58 18 North Carolina (21) 39 12 Texas (7) 25 2 Utah (9) 45 4 Washington (12) 44 1 Wisconsin (12) 33 5 Unassigned 14 TOTAL 478 b 93 a Number in parentheses following field center is number of analysts on staff. b Total does not equal 232 because some industries were assigned to more than one field center. SOURCE: Booz, Allen & Hamilton, Inc. (1979: Exhibit IV-3).
118 WORK, JOBS, AND OCCUPATIONS only one industry. Since these three field centers have approximately the same number of analysts, the reason for this apparently uneven distribu- tion of responsibility is unclear. ESTABLISHMENT SELECTION Once field centers received their assignments, they independently deter- mined how best to fulfill them. Although the Handbook for Analyzing Jobs (U.S. Department of Labor, 1972) provides relatively detailed information on how to conduct a job analysis study within a firm, it contains virtually no guidance on how to go about studying an industry per se or on how to select firms within an industry. In most field centers each analyst was given sole or lead responsibility for a set of industries. The number assigned per analyst varied by the size of the industries involved and by other factors such as the size of the field center and the extent of its involvement in activities other than job analysis (see chapter 5 for a discussion of field center organization and activities). Analysts pursued their assignments relatively autonomously. There are indications that field centers varied in the degree of supervision or direction given to analysts. Several centers instituted informal procedures whereby analysts submitted a general study design to the center supervisor or supervisory analyst before proceeding, but this was not required by the national office. Analysts typically began by doing background research on the industry in question in order to acquaint themselves with its processes, products, and jobs. This usually involved library research and reference to the appropriate trade publications, the Standard Industrial Classification (U.S. Department of Commerce, 1972), the Occupational Outlook Hand- book (U.S. Department of Labor, 1978a), and others. Some analysts also consulted old job analysis schedules and occupational code requests as background material. The Standard Industrial Classification appears to have been especially useful at this point. Because it is much more detailed than the DOT'S industry designations, it provided a means by which analysts could determine an industry's subgroupings and hence develop a framework for pursuing their assignment. As soon as analysts were satisfied that they were familiar with the industry and the types of jobs found in it, they selected establishments that seemed likely to have jobs typical of the industry. In locating establish- ments, analysts relied on various sources: industrial registers (e.g., Thomas; Dun & Bradstreet), the classified sections of telephone directo- ries, ES-202 forms (submitted to state Employment Service offices by companies that contribute to unemployment insurance), directories of
Procedures Used to Produce the Fourth Edition' DOT ~9 trade and professional associations, indexes of local manufacturers, and publications of the Better Business Bureau and the Chamber of Com- merce. Many of these publications contain a good deal of information on the firms available for study in the field center's area, often listing the firm's sac code, address, the names of company officers, the number of employees (sometimes disaggregated by sex), and the products manufac- tured. Despite the absence of guidelines, there was substantial uniformity among the field centers on the criteria used for the selection of establishments. The primary criterion appears to have been proximity to the center. For some field centers, proximity meant the larger federal region in which they were located; others stayed within their state borders (those with state-imposed restrictions on out-of-state travel); and others stayed within the immediate metropolitan area. One field center supervisor reported, for example, that analysts traveled to firms outside the city in which the field center was located "only as a last resort." Overall, the bias appears to have been toward staying as close as possible to the field center in fulfilling assignments. Analysts attempted to select at least one small, one medium, and one large establishment within an industry or (for large industries) subindus- try. Analysts repeatedly expressed the opinion that size was an especially important source of variation in jobs. In their experience, what was a single job in a small company would often be broken up into several jobs in a large firm. Size was apparently assessed by referring to information on the relative number of employees in various establishments in the area. Although this assessment appears to have been a rather casual one, in at least one field center the supervisor reported that the size distribution of local establishments in a particular industry was obtained from locally available publications and divided into thirds so that one or two establishments could be selected for study within each third. Analysts also tried to select establishments that they believed might be employing new or emerging technologies, on the assumption that new jobs would be available for job analysis. Having completed background research on the industry in question and having selected several possible establishments for study, analysts then attempted to gain the employer's consent to go on site for job analysis. Thus availability, predicated on employers' consent, was the final criterion of establishment selection. Various approaches were used to gain access. The supervisor of one field center reported having the local Employment Service office set up the initial meeting between the analyst and company personnel. At most field centers, however, analysts contacted the appropri
120 WORK, JOBS, AND OCCUPATIONS ate company officer and arranged for the study on their own. Without exception, analysts reported that they tried to talk to the most senior person in the company and were more successful in gaining access when they did so. To obtain an employer's consent, analysts used a "sales pitch," emphasizing the benefits to the employer of the data to be collected. In unionized companies, union consent sometimes was required, although it was usually sufficient merely to notify unions when analysts were to be on site. It was difficult to determine the degree to which or the circumstances under which employers cooperated or refused to cooperate in a study. Although at least one field center maintained a file of all establishment contacts and their outcome (including, when applicable, the reason for refusal), field centers were not required to do so. As a result, there appears to be no systematic way of assessing the extent, nature, or implications of employers' noncooperation. We received mixed reports about the rate of refusal. There appears to have been a good deal of variation across field centers. No clear picture emerged as to the type of company most likely to have refused. Analysts at two field centers reported that small companies were more likely to do so and attributed this to the independent, "get-off-my-back" style of many such firms and to their general antipathy toward government intervention of any sort. At another field center, however, analysts could discern no pattern to nonresponse, saying that small companies were no more likely to refuse than large ones. There was some feeling that large companies were more cooperative than small ones. Gaining access to large companies often entailed extensive negotiations and permission from the company's home office (often located in another state or region), but several analysts said that if they persevered, they were usually successful. JOB ANALYSIS PROCEDURES Before beginning an establishment study, analysts discussed the proposed study with the employer or supervisory personnel, explaining its purpose and intent, going over procedures, and learning the physical layout of the firm. Even when employers agreed to allow a study to be conducted, they sometimes imposed certain restrictions on analysts' activities. It was reported, for example, that employers (and unions) typically did not permit analysts to use tape recorders or decibel meters in the analysis of production jobs. Some jobs were declared ok limits for security reasons, usually to protect a unique aspect of a manufacturer's production process.
Procedures Used to Produce the Fourth Edition DOT STAFFING SCHEDULE AND ORGANIZATION AND PROCESS FLOW CHARTS 121 In the next step of a typical study, analysts completed a form called a staffing schedule, a copy of which is shown in Figure 6-1. Instructions for completing the staffing schedule are laid out in some detail in the Handbook for Analyzing Jobs (U.S. Department of Labor (1972), hereafter referred to as the Handbook). Schedules were completed for every division of an establishment if it was analyzed in its entirety or, if only a portion was analyzed, for the relevant divisions or departments. The staffing schedule was usually prepared with the assistance of the establishment's personnel office or supervisor. If a company had a personnel office, analysts usually copied its records verbatim to complete portions of the staffing schedule. Otherwise, they compiled the necessary data anew. The staffing schedule has two parts, a face sheet and a title sheet. On the face sheet, analysts entered identifying information about the establishment (a unique control number assigned by the field center and a Standard Industrial Classification code) and names of products manufac- tured or services rendered. Analysts then completed one or more title sheets. To do so, they listed company job titles by organizational units for the entire establishment or for the divisions in which they were interested. For each job the total number of workers was recorded, as were subtotals broken down by sex and by shift, if applicable. If the job was entry level, this was also noted. Although it is nowhere stated in the Handbook, it was reported to us that part-time and trainee jobs were not listed on the title sheet and hence were excluded from study. On the basis of a preliminary inspection during a plant tour and/or in consultation with management or other personnel, analysts then deter- mined for each job whether it was substantially similar in basic tasks and requirements to a job definition in the third edition DOT. If so, a third edition title and preliminary code were entered in the appropriate space on the staffing schedule. If a job was a combination of two or more DOT occupations, all the applicable titles and codes were entered; if a job could not be converted to a third edition code, the space was left blank. This comparison of jobs in an establishment against third edition titles determined the scope and direction of the ensuing study. Jobs that could be converted to a DOT definition but were not specific to the industry being studied (e.g., clerical jobs in manufacturing industries) were not supposed to be analyzed. Instructions in the Handbook are ambiguous about whether all remaining jobs are supposed to be analyzed. It appears that they should be, but departures from complete top-to-botton studies were not uncommon.
122 Cq To P: o ~ o ..Q ·- o ~ ~ . ~¢ c3 ~ B Ha. ~ _ ~ it, ¢ Em - ¢ o C) C. of to o <D Cat Cal .,' v) V] .,1 Q) Ha o 4~ o .. CO U] o V ~^ ~_ :3 CO o ._ I., 02 ~ ~I o ~1 ~1 cd aJ o s o ¢ to CO o o U] $~ :5 c) ·,, s o' o h q~ C bO .,' · - U] ~ o ~D 00 C~ a) h ao o Vl ~n cC 4~ u] Ul ·~. s E~ . ~ p: ~q .O o s~ c) c oD ~l u] o s ~n eJ .~' ~Q OD c7 a~ Ul h o q~ U] G) ~n ba eJ 00 ~D o U] tD U7 U] C) o Ul o o C, 4~ o U' N 4 ~ h .. (D ~Ul J4 Ul ~D C ~C.) ~ o ¢
123 <- Cq~ ~, ~ 1 of m1 ° _ ~ a cat a En v A ~ _ ~ $ ¢ En o9 ¢ En o l ~1 b b U] 4 ._ .= ¢ ~¢ O ~EM ~ O . . - o o 4) ho .St . . C) o V, be: C) I tl~ Cat -0 ' 00 1 ~ 00 rid I ? ~1 ~ E,3~s ~ _ .- =3 . C.) O ~ O ~ ~ .~ O ~ E-l ~ I ~ E' ~ o E E.b ~ 1 lc o - - - - - - c~ o c~ l bO o 1~ ~ oo C2 cD bO o o CD bO CD oo c. o, o o ~D bO =: 1 1 c~ o, bO uo) o CD ~: ¢ ~ x :^ 3 o o ~9 bO ~_ ~o a, o ·~ ~q (~.) 0 =, c,, ao i~ ~ 3 o o O r9 Q ~ _ ~ 1 ~ 1~ ~ ~o X ~C , 0~^ ma ~ ~E°° ~0 ~ ~ ~ a ~ o D ~q~ m~ ~ 0 ~u, * oo co c~ 0 u: 0 CD bO 1 e' , ~ 1 1 1 ~ 1 -1 oo 1 C~ 1 ao · 1 ~ o 1 o ao 1 ~ CD 1 '9 1 a 1 - 1 1 ~ x a 4 - ~r; a E~ o o~ C2 CD c~ o oo N z Ix~ O ~ ~ j a ~; ~ | Z ~ 9; ~ 1 ' ~ " . . 1 1- - 1 1 1 1 1~ 1 . 1 1 4> 1 1 1 ~1 1 1 ~1 1 1 ~1 1 1 ~1 1 1 o 1 1 .~ 1 1 o 1 ~ 1 tn 1 1 cn . 1 ~ 1 ~ 1 C~ 1 ~ 1 I , A 1 i 1 ~ 1 1 ° 1 rm 1 1 -1 1 ~ i 1 ~ 1 1 ~ 1 1 1 Iz 1 1 ~ 1 1 ~ 1 I o . 1 1 ~ 5 1 1 ~c ~ 1 =: 1 1 ~ ~ 1 1 Z cC 1 1 ~- 1 1 1 1~ 1 1 1 i ~ i 1 ~ 1 1 ~ 1 1 ° ! a) · 1 ~ I Z o ~ ~ ~Z i i | a 1 1 1 CD 1 ~ 1 ~ 1- i' ~1 1 ~ 1 o 1 a) 1 a' a) 1 1 · 1 1 oo 1 ° 1 '9 1 (D 1 1 ~o 1 ~ =0 I h 1 ~ g I s0 1 ~ 3z 1 0 1 ~ 1 1 - 1 io - ~C a' C) 0 OD `,, s - C~ 1- 1 1 ~ 1 1 ° 1 1 4~ 1 1 a, 1 1 ~ 1 1 4) ~ 1 ~ 1 1° 1 1 a) 1 1 C 1 1= 1 1 ~ 1 1 ca 1 1~ 1 1 ~ 1 I 0 I 1 " 1 1 ~ 1 I. 1 1 ~ 1 c' 1 bO 1 ~ 1 ·,1 1 1 1 _' 1 a:` 1 1 ^ o UO bO Qo ~
124 WORK, JOBS, AND OCCUPATIONS For jobs that could not be converted to a third edition title, a complete job analysis was undertaken. A slightly abbreviated analysis was made of jobs that could be converted as well as of jobs that could not be converted but that the analyst knew had already been analyzed in other establish- ments. At a later date, after they had finished analyzing jobs in the establishment, analysts entered "treatment codes" for the remaining jobs to indicate how thoroughly each had been studied. In the final step before beginning the intensive analysis of particular jobs, analysts diagramed the company's organization arrangements, indicating lines of authority between different divisions. They also prepared a "process flow chart" to depict the "sequence of procedures or processes at the establishment: (a) if industrial, from arrival of raw materials to shipment of finished product; (b) if service, from entry of client or material into the service until the service is completed on the individual or item" (Handbook, p. 55~. Process flow charts were usually prepared only for company divisions that were directly engaged in production and not for supporting departments such as quality control, sales, accounting, etc. These departments, however, were included in the organization chart. The Handbook offers very little instruction about the completion of either the organization or the process flow charts, and it seems that frequently these charts were not prepared. JOB ANALYSIS All of the information obtained for the staffing schedules, organization arrangements, and process flow charts was intended to assist analysts in identifying jobs for study and in keeping track of their progress as the establishment study proceeded. The Handbook (p. 3) defines "job" as "a group of positions which are identical with respect to their major or significant tasks and sufficiently alike to justify their being covered by a single analysis. There may be one or many persons employed in the same job." The various components of a job according to the methodology used in the occupational analysis program are elements, tasks, and positions. "Element" is "the smallest step into which it is practicable to subdivide any work activity"; "task" is "one or more elements" that form a distinct activity or step in the performance of work; and "position" is "a collection of tasks constituting the total work assignment of a single worker" (Handbook, p. 3~. Thus there is a one-to-one correspondence between a worker and a position. Despite these conceptual distinctions between a job and its component parts, analysts generally accepted the establishment's definition of which positions constituted a job. Job analysis for the DOT iS based on a combination of observation and
Procedures Used to produce the Fourth Edition DOT 12S interviews of workers performing a particular job. Although analysts usually asked to observe average workers, employers frequently directed them toward their best workers. When this occurred, most analysts tried to observe other workers in the job. Some analysts, however, preferred to observe and interview better workers, who were thought to be more knowledgeable and articulate about their jobs. Analysts usually observed one or two workers for each job they analyzed. The conduct of the observation and interviews was structured to some extent by the requisites of the job analysis schedule, which is discussed in more detail later in this chapter. With an eye to completing the schedule, analysts noted the tasks entailed in the job and the percentage of time spent on each; the machines, tools, or work aids used; working conditions; and a variety of other information. Analysts either recorded data directly onto the job analysis schedule, recorded it on worksheets developed by the local field center, or simply took notes that they later transcribed. One analyst estimated that about 65 percent of the analyst's time was spent interviewing workers or their supervisors and the rest in observing and taking notes on the conditions of work, work aids, products produced, etc. Job data obtained from observing and interviewing workers were usually checked with the workers' immediate supervisor. In some situations the analyst was unable to talk directly with the worker. For production work this most frequently occurred when workers were engaged in complex or protracted tasks, on an assembly line in which the pace had to be maintained, doing piecework, and/or in noisy surroundings. For other kinds of jobs this limitation arose when workers were involved in a wide variety of tasks over long periods of time. The Handbook offers no clear-cut alternative procedures for obtaining data about jobs in such situations, advising only (p. 14) "~1) using establishment job descriptions or specifications supplemented by discussions with administrative and technical personnel; (2) obtaining job descriptions, specifications, hiring requirements, and related data for certain jobs from associations, societies, and other similar organizations; or (3) interviewing workers, supervisors, and/or management in a place apart from the work site." In an effort to deal with this problem for clerical jobs, one field center developed a special questionnaire to ascertain job tasks and requirements. The questionnaire was completed by the worker and followed up by analysts with a brief interview. These data collection procedures were repeated for each job in the establishment for which an analysis was needed. Depending on the size of the establishment and the number of jobs to be analyzed, the on-site analysis portion of an establishment study took between 1 week and 2 months to complete. Usually, analysts had to work fairly rapidly so as not
126 WORK, JOBS, AND OCCUPATIONS to disrupt the company's work routine. Once they completed their on-site analysis, it was usually difficult to return to the establishment for a second round of observations. Writing the Job Description' and Assigning a DOT Code Having completed the on-site establishment study and with data in hand for a number of jobs, analysts reviewed their work before filling out a job analysis schedule for each job. Completion of the job analysis schedule, an example of which is given in Figure 6-2, required analysts to supply information about the job (e.g., establishment job title, sac code and title), to write a job summary, and to code and/or rate much of the information they had gathered. The job summary provides information on the purpose and nature of the job and on the level and significance of the worker's involvement with data, people, and things. The analyst also listed licenses or certification required for the job; noted the job's relationship to other jobs in the firm; listed and described (if relevant) machines, tools, equipment, and work aids used by the worker and materials and products with which the worker was involved; provided a detailed description of job tasks (including percentage of time devoted to each); and defined teens used in the schedule that could not be found in a standard dictionary. If a particular job could be converted to a third edition DOT code (treatment type C) or was similar to a job for which the analyst knew a schedule had already been prepared (treatment type V), the detailed description of job tasks was not provided. Instead, analysts listed only significant differences between the job being analyzed and the DOT definition or original job description. In light of the above information the analyst assigned numerical codes (available from the Handbook) for each of the following: (1) one orimore of the 100 work fields (e.g., hunting-fishing, butchering, research), Which relate to "specific methodology~ies) used in the job-worker situation" (Handbook, p. 27), (2) one or more materials, products, subject matter, or services (MPSMS) (e.g., lumber and wood products, business services and administration) that together with the work fielders) reflect "the specific technology with which the worker is involved" (Handbook, p. 27), (3) worker functions measuring the complexity in relation to data, people, and things. On the basis of job tasks, analysts indicated whether the relationship was significant according to the highest level of worker's involvement on each of the three dimensions. To assist analysts in completing this portion of the schedule, the Handbook provides descrip- tive examples of each level. On the basis of these ratings and the job summary the analyst then
Procedures Used to Produce the Fourth Edition DOT U.S. Department of Labor Manpower Administration JOB ANALYSIS SCHEDULE 1. Estab. Job Title INFORMATION DESK CLERK, receptionist-clerk 2. Ind. Assign. ret. tr. 3. SIC Code(s) andTitle(s) 5311 Department Stores co u, . cot to c). ' ICcDg a .,. 0 ~ c, 0 0 A; a 0~ ED ca :^ oh a ~u, a u, ~5,4 .- ~a, O ~ 127 OMB 44-R0722 Estab. & Sched. No. 071-3120-423 4. JOB SUMMARY: Answers inquiries and gives directions to customers, author- izes cashing of customers' checks, records and returns lost charge cards, sorts and reviews new credit applications, and requisitions supplies, working at Information Desk in depart- ment store Credit Office. 5. WORK PERFORMED RATINGS: (D ) (P) T 1 1 ~ Worker Functions ~ | People | Things | 1 5 1 6 1 7 1 Work Field 282-Intormation Giving 231 Recording M.P.S.M.S. Qqn-R~siness Service 6. WORKER TRAITS RATINGS: GED 1 2 (3) 4 5 6 SVP 1 2 (3 ) 4 5 6 7 8 9 Aptitudes G 3 V 3 N 3 S 4 P 4 Q 3 K 4 F 3 M 4 E 5 C 5 Temperaments D F I J (M)(P) R S T (V) Interests la (lb) (2a) 2b 3a 3b 4a 4b 5a 5b Phys. Demands (S) L M H V 2 3 (4) (5) 6 Environ, Cond. (I) O B 2 3 4 5 6 7 MA 7-36 FIGURE 6-2 Job analysis schedule. Source: Handbook for Analyzing Jobs (U.S. Depart- ment of Labor, 1972:37-41). Continued overleaf
128 7. WORK, JOBS, AND OCCUPATIONS General Education a. Elementary 6 High School none Courses b. College none Courses 8. Vocational Preparation a. College none Courses b. Vocational Education none Courses c. Apprenticeship none d. Inplant Training none e. On-the-Job Training 3 to 5 weeks by Credit Interviewer f. Performance on Other Jobs none Experience none 10. Orientation 1 week 11. Licenses, etc. none 12. Relation to Other Jobs and Workers Promotion : From this is an entry Job Transfers: From none Supervision Received CREDIT MANAGER Supervision Given none To CREDIT INTERVIEWER To none 13. Machines, Tools, Equipment, and Work Aids Impressing Device - Small Hand-operated device, of similar construction to stapler with a nonmoving base and a moveable upper arm (cont'd) 14. Materials and Products none FIGURE 6-2 Continued
Procedures Used to Produce the Fourth Edition DOT 15. Description of Tasks: 129 Answers inquiries and gives direction to customers: Greets customers at Information Desk and ascertains reason for visit to Credit Office. Sends customer to Credit Interviewer to open credit account, to Cashier to pay bills, to Adjustment Department to obtain correction of error in billing. Directs customer to other store departments on request, referring to store directory. (50%) 2. Authorizes cashing of checks: Authorizes cashing of personal or payroll checks (up to a specified amount) by customers desiring to make payment on credit account. Requests identification, such as driver's license or charge card, from customers, and examines check to verify date, amount, signature, and endorsement. Initials check, and sends customer to Cashier. Refers customer presenting Stale Date Check to bank. (570) 3. Performs routine clerical tasks in the processing of mailed change of address requests: Fills out Change of Address form, based on custo- mer's letter, and submits to Head Authorizer for processing. Files customer's letter. Contacts customer to obtain delivery address if omitted from letter. (10%) Answers telephone calls from customers reporting lost or stolen charge cards and arranges details of cancellation of former card and replace- ment: Obtains all possible details from customer regarding lost or stolen card, and requests letter of confirmation. Notifies Authorizer immediately to prevent fraudulent use of missing card. Orders replacement card for customer when confirming letter is received. (10%) 5. Records charge cards which have inadvertently been left in sales departments and returns them to customer: Stamps imprint of card on sheet of paper, using Imprinting Device. Dates sheet and retains for own records. Fills out form, posting data such as customer's name and address and date card was returned, and submits to Authorizer. FIGURE 6-2 Continued Continued overleaf
130 WORK, JOBS, AND OCCUPATIONS Makes impression of card on face of envelope, inserts card in envelope, and mails to customer. (5~) 6. Sorts and records new credit applications daily: Separates regular Charge Account applications from Budget Accounts. Breaks down Charge Account applications into local and out-of-town applications and arranges applications alphabetically within groups. Counts number of applications in each group and records in Daily Record Book. Binds each group of applications with rubber band, and transmits to Tabu- lating Room. tlO%) 7. Prepares requisitions and stores supplies: Copies amounts of supplies requested by Credit Department personnel onto requisition forms. Submits forms to Purchasing Officer or Supply Room. Receives supplies and places them on shelves in department store storeroom. (10%) 16. Definition of Terns Stale Date Checks - More than 30 days old 17. General Comments none 18. Analyst A. Yessarian Date 7/25/70 Editor M. Major Date 7/26/70 Reviewed By John Milton National Office Reviewer W. Irving Title, Org. Credit Manager FIGURE 6-2 Continued
Procedures Used to produce the Fourth Edition DOT SUPPLEMENTAL SHEET Item 13 (con.) containing inked rollers which Impressing Device (con) are moved by a lever in the upper arm. Charge card is placed in a grove in the base, stand-up print facing up, and paper or bill positioned over card, then the upper arm is brought down and lever depressed to bring inked rollers over paper to make impress of card's print. FIGURE 6-2 Continued 131
132 WORK, JOBS, AND OCCUPATIONS TABLE 6-2 Worker Trait Summary Worker Trait Number of Factors T. . . raining time Aptitudes Temperaments Interests Physical demands Environmental conditions TOTAL 4 11 10 s 6 7 43 assigned the job its DOT code: the first three digits were assigned primarily on the basis of work fields and MPSMS; the next three digits are the worker function ratings of actual level (rather than significance) of involvement. If the job was identical in all significant respects to a published definition, the analyst also entered a DOT title and industry designation for it. Rating Worker Traits Another step in completing the job analysis schedule entailed rating jobs for the six types of worker traits discussed in chapter 2. In all, this involved assigning scores on 43 distinct factors, as shown in Table 6-2. The rating task was by no means straightforward, requiring detailed knowledge about the job as well as a great deal of judgment on the part of the analyst, since many of the factors were not easily observed or measured. To assist analysts in rating worker traits, the Handbook contains numerous illustrative situations or bench marks for each level of each factor of each trait. Analysts' opinions of the usefulness of these bench marks were mixed, however. Some analysts did not find them comprehensive enough; some found them inconsistent, contradictory, and confusing; and some apparently felt that they were adequate. The Handbook also provides instructions as to how and on what basis each worker trait should be assigned. Both these instructions and the definitions of the traits are brief and vague, consisting usually of only one or two paragraphs. Each job was evaluated for three categories of GED (reasoning, math, and language), for which the scale is given in Figure 6-3. After determining the level required in each category the analyst selected the highest level of the three to express the final GED for the job. For svP, analysts assigned one of nine levels, which correspond to ranges of training time but do not reflect the type of training or where it has been acquired
Procedures Used to Produce the Fourth Edition DOT 133 (svP rating levels are shown in Figure 6-4~. The svP was reported to be difficult to rate because the frame of reference for measuring the amount of training was unclear. Training times were determined by considering a variety of data collected during the on-site study: employer's hiring requirements, union specifications, workers' qualifications, types of work aids used (e.g., calculators, gauges, etc.), and types of tasks performed (e.g., arithmetic calculations, writing, etc.~. The Handbook cautions analysts not to rely too heavily on the qualifications demanded by the employer or union or on those that workers bring to the job but rather to assign GED and svP primarily on the basis of skills or tasks intrinsic to job performance. According to most analysts' reports, however, employers' hiring require- ments figured prominently in the assignment of these ratings, especially svP. The analyst next rated jobs for 11 aptitudes indicating the level of each aptitude required for "average, satisfactory performance" (~Handbook, p. 233~. A single rating scheme, shown in Figure 6-5, is used for all aptitudes. Scores of 1 (high) to 4 (low) were assigned to indicate the level required for satisfactory performance; 5 was assigned if the aptitude was not required to perform the job. Level 5 could not be assigned on the intelligence aptitude (G), however, since it was "assumed that every job requires at least a '4' level of this aptitude" (Handbook, p. 294~. Temperaments were rated next. The temperament component consists of 10 factors, shown in Figure 6-6. In rating temperaments the analyst selected those that were "important in relation to the kinds of adjustments which the worker must make for successful job performance" (Handbook, p. 3134. The presence or absence of a given temperament, rather than the level or degree required, was indicated. The analyst next evaluated jobs for each of five pairs of bipolar interests, selecting those that were important for job performance, thereby automatically rejecting the other of the bipolar pair. The interest factors are shown in Figure 6-7. The aptitude, temperament, and interest traits were generally perceived to be more ambiguous than the others. Analysts could offer no ready explanation for how they decided on these ratings, other than to say that they learned through experience or that it was a matter of getting a feel for the job. Aptitude ratings were evidently especially difficult. A number of analysts said that they would have felt much more confident about indicating the presence or absence of an aptitude rather than the degree required. Since for the temperaments and interests an indication of simple presence or absence was all that was necessary, these ratings appear to have posed fewer problems. Finally, jobs were rated for each of the following six physical demands
134 ~ R ~-- it--~ ~ ~ R ~ _ ° ~ ~ P ~ ~ ~ p O ~ P ~ ~ _ h ~ E gg E i! a ~ ha go a 8 a E=' En ~ ;, a ~ h ~ ~ a a ~ E] ¢ ~ 0 ¢ ¢ E" no V ¢ Fl o o ¢ A: 1 Ed ,a ED Ding ~= -c ON,, ~2 R ~ cd use ~ P 4~ ~ b ~ ~ ~ E id, ,~, rJ ~, ~ ~ ~ e ~ ~ ,, be ,£ P .= P .0 . co A 4) 0 cot u, E ~ 8 ' ~.~r 8 ~5 j~§ , . E ~ ~ p ·-. ° ~ p. ~ ~ ~ b ~ ¢ ~ E ~ g 8i . `,~ B p b S . o ~ ~ ~ ~ ~ ¢ c: ~ u, ^= u~ R 4~-- ~ t_ ·~=o~ g~ P ~ £ ,._ E E Y 1: ~ ~ ~ ~ oo t~C ~ ~ ~ _ o ~ ~ ~ ~ ~ ~ 8 o o ~ O or, ~ b - 0 ~ ~ ~ ~ x e ~ .,'. ·b B ° `4 R b 05 b ~= tR b b 'S So U:
135 .~t Pi ~ E to o 2 ~ ~ .° ~ ~ ~ b o ~ ~ ~ p ~a ]1~3i~=i53~- o o ~ i I , in let P: ~ ~ ~ A& ¢ ~ E b ~·o ~ ~ h ~ p ~ O ~ . ~ 8 ~ O .^ ~- O E of. ~ , ¢ ~ °O W V) o ~ ~ _ ~ cq~ ·5 .--~ ~ fi ,~ fi ~ ~. o Bolt ^= 4,, Pi the.-~ ~ ~ ~ ~ 3 ~ ah pa ~ O ~ ~ ~ <at O ~ ~ ~ ~ O h W fi o ~ o,~, ~ ~ o a 3 ~ ,~, b ~ ~ ~) 5° ~ ~ · b o ~ a fi ~.~ O' bq-- b .8 fio .o fi ~ b ^._ --_ = = 0 ~ ¢ O O .= ~ - C~ ~DO o U. .D . . C> o V - .~ o ~' o .~ 00 CO V, CO . ~
136 ED Y o a V ¢ ¢ o 0 = O b b 3 59.0-_ ~ ~= ~ ~ _ Q0 0 ~ O D X ~ ~3 ~ ~ o 3 b ~ ~ ~e ~ b C) >6,= ~ A, O ~ ·~,4 b bo o ~ bit .~0 C , o ~ C P bt 0._ be.? bC ~ ~ ~ ~ ~ · - ~ ~ ~ b40 O O p: ~ . - ~ ~ O ~ ~ =-~ P: ~ ~°- Fib == U] ~ o ~ ~ o , ° ~ a 3 a) lo. _ It ~ CO a) ~0 U] ~C: ·= O o ~ ~ . - . c~ =3 ~. ~ = o ~iV bL~ · . - ~ - ~q .= ~ ~ ~ ~ - = ~ ~ ~ ~ s-~= ~ o ~== '-~ ~ 3 E ~ ~E E 1 ~ ~ ~= C. ° ~ ~ ~ h O U] o 0 ~ ~ b4~ o .~.- ~ ~ ~ 3 ~ ~o b >- ~ ~ =P~p~. o-C-OcI)-iq _ ~ E 2 ~ 1 b] ._ C~ E~ W Y o W ¢ V E~ ¢ E~ ¢ W o~ W o ¢m P: _, _, Ct ~ c! ~V .= =. ~ O ~D,- ~ 2 2 S ~ y O '- S O co .0 ~ b _ U} ~ ~ · O ¢ - ~ Ul · . ~ ~ .S a ~q o . e; S~ U ~CtS ~Q O ·O- 60 ~ O 4~ ~ ~ ~ ~ ~ b4 aS °,c,=: =_m {,, 4~. - e~ O ~U: ~ · ~ ~ ~ C ~ ~ - -> b ~ O ~ O D ~ C) b. = ~ co ~ C _ a' ~ ~ ~ -3 ~ ~ ~ b == _~ O ¢~-~ ~ ~ b ~ o O o O._ O .0 aq O ~ U) 0 1- V: 41) o o ~ =0 O .E 3 ~.o ~ sq ~ ~ - _ o c5 =, - aS u] _ ~ C0._ O ~ c, en o N .o .~ > b ._ t0 ._ ~Q g fi ~ >` ~ . B .u .~ q} .~ 8 . - . .. b o - rl ~_
Procedures Used to Produce the Fourth Edition DOT Level Time ~ 1 Short demonstration only 2 Anything beyond short demonstration up to and including 30 days 3 Offer 30 days up to and including 3 months Over 3 months up to and including 6 months 5 Over 6 months up to and including 1 year 6 Over 1 years 7 Over 2 years 8 Over 4 years up to and including 10 years 9 Over 10 years year up to and including 2 years up to and including 4 - ~ Time spent in s~eneral educationa1 develc~pmer.r is not considered in estimating specific vocational preparation, 137 FIGURE 6-4 Scale for specific vocational preparation (svP). Source: Handbook for Analyzing Jobs (U.S. Department of Labor, 1972:220). and seven environmental conditions. Physical demands were (1) strength (lifting, carrying, pushing, and/or pulling), (2) climbing and/or balancing, (3) stooping, kneeling, crouching, and/or crawling, (4) reaching, handling, fingering, and/or feeling, (5) talking and/or hearing, and (6) seeing. Environmental conditions were (1) work location, (2) extreme cold with or without temperature changes, (3) extreme heat with or without tempera- ture changes, (4) wetness and/or humidity, (5) noise and/or vibration, (6) hazards, and (7) atmospheric conditions. The analyst indicated simply the presence or absence of physical demands 2 through 6 and environmental conditions 2 through 7. Strength was rated according to one of five levels to reflect sedentary, light, medium, heavy, or very heavy work. Work location was rated to indicate whether the job was performed primarily indoors, outdoors, or both. Many of these factors could have been measured objectively. Because analysts did not take thermometers, decibel meters, or other instruments and gauges with them on site in fact, as was mentioned previously, they
138 Intelligence V Verbal N Numerical S Spatial P Form Perception Q Clerical Perception Quintiles for Rating Aptitudes WORK, JOBS, AND OCCUPATIONS K Motor Coordination F Finger Dexterity M Manual Dexterity E Eye-Hand-Foot Coordination C Color Discrimination 1. The top 10 percent of the population. This segment of the population possesses an extremely high degree of the aptitude. 2. The highest third exclusive of the top I O percent of the population. This segment of the population possesses an above average of high degree of the aptitude. 3. The middle third of the population. This segment of the population pos- sesses a medium degree of the aptitude, ranging from slightly below to slightly above average. 4. The lowest third exclusive of the bottom 10 percent of the population. This segment of the population possesses a below average or low degree of the aptitude. 5. The lowest 10 percent of the population. This segment of the population possesses a negligible degree of the aptitude. FIGURE 6-5 Aptitude factors and rating scale. Source: Handbook for Analyzing Jobs (U.S. Department of Labor, 1972:233). p R S D pop (direction, control, and planning) F FIF (feelings, ideas, or facts) I INFLU (influencing) J SJC (sensory or judgmental criteria) M MVC (measurable or verifiable criteria) DEPL (dealing with people) REPCON (repetitive, continuous) PUS (performing under stress) STS (set limits, tolerances, or standards) VARCH (vanety and change) FIGURE 6-6 Temperament factors. Source: Handbook for Analyzing Jobs (U.S. Department of Labor, 1972:29S).
Procedures Used to Produce the Fourth Edition DOT 139 la. A preference for activities vs. lb. A preference for activities dealing with things and concerned with the com objects. mun~cation of data. 2a. A preference for activities vs. 2b. A preference for activities involving business contact of a scientific and technical with people. nature. 3a. A preference for activities vs. 3b. A preference for activities of a routine, concrete, of an abstract and creative organized nature. nature. 4a. A preference for working vs. 4b. A preference for activities for the presumed good that are carried on In rela of people. lion to processes, machines, and techniques. Sa. A preference for activities vs. 5b. A preference for activities resulting in prestige or the resulting In tangible, pro esteem of others. ductive satisfaction. FIGURE 6-7 Interest factors. Source: Handbook for Analyzing Jobs (U.S. Department of Labor, 1972:317). were often prohibited by employers from doing These ratings too were based on analysts' judgments. The analyst queried workers closely about the processes, machines, and materials they worked with in order to determine environmental conditions. To assess physical demands, job tasks were usually merely observed. COMPLETING AN ESTABLISHMENT STUDY After completing the establishment staffing schedule, organization and process flow charts, and job analysis schedules for each of the jobs being analyzed in the establishment, the analyst summarized the study in a narrative report. As its title implies, the narrative report is essentially descriptive. Although there is no prescribed format for writing these reports, the Handbook (p. 59) suggests that they be organized to include information on (1) the purpose and history of the establishment and the scope of the job analysis study, (2) environmental conditions, (3) the organization and operations or activities of the firm, and (4) its personnel policies and practices. As the study proceeded or at its completion, the analyst submitted the materials produced to either a lead analyst or the field center supervisor for review. Once the study had been checked and approved, the analyst often sent a note of appreciation to the employer and, depending on the
140 WORK, JOBS, AND OCCUPATIONS field center's practice, either gave the employer a copy of all materials or provided copies on request. MODIFICATIONS OF PROCEDURES The steps described above for conducting an establishment study were in effect during the normal course of fourth edition production and reflect standard operating procedures. During certain periods of production, however, and sometimes in the normal course of events these procedures were modified. Several years prior to the anticipated publication of the fourth edition, national office staff decided that in order to increase coverage and expedite the production process, modifications to traditional operating procedures were necessary. From 1974 to 1976, analysts were directed to concentrate their efforts on verifying jobs against existing job schedules for similar jobs in other establishments or against the DOT definition if the job could be converted to a third edition code. In this way much of the time-consuming writing entailed in completing the job analysis schedule was eliminated. Evidently, this directive was variously interpreted by the field centers. Analysts at some field centers continued to produce job analysis schedules according to Handbook procedures, which require a complete study of the job being verified; others resorted to shortcuts, telephoning an establish- ment or trade association, for example, in order to verify descriptions of jobs being analyzed in other establishments or to confirm a third edition definition. When this procedure was followed, staffing schedules were not produced. At about the same time that the directive to change the standard procedure was issued, field centers were also asked to review the status of their industry assignments. As part of this review, field centers submitted lists of jobs (primarily unanalyzed third edition jobs) in their assigned industries that they had been unable to analyze. All such jobs were compiled in a so-called Not Available and Obsolete (NA&O) list that was then circulated among the field centers. Field center staff were requested to try to locate and analyze those jobs on the list that were available in their region; to do so, a complete establishment study was not required. Thus most of the jobs on the list were picked up piecemeal, analysts often entering an establishment to analyze one or two jobs rather than the entire company or a division within it. Again, in this circumstance, staffing schedules were rarely produced. These changes in procedure verification and NA&~were instituted under the pressure of an approaching deadline. Many analysts reported that these pressures resulted in a drop in the quality of the studies and schedules produced during this period. The modifications, however,
Procedures Used to produce the Fourth Edition DOT TABLE 6-3 Definition Writing Assignments Percentage Occupation of Base Center Category Titles California, national office 0-1 New York Florida Washington State North Carolina TOTAL 4 5-9 12 74 100 141 appear to have had the intended effect of increasing the quantity of jobs analyzed: Booz, Allen & Hamilton, Inc. (1979) reports that 30 percent of the schedules supporting the fourth edition were produced during the 1974-1976 period in which these procedures were in effect. Finally, no attempt was made to observe certain types of jobs, including some professional jobs, seasonal jobs, and jobs involving a wide variety of tasks spread over long periods of time. The methodology used by the occupational analysis program, because it relies heavily on the direct observation of jobs to collect information about them, is not feasible for the analysis of such jobs on site. Instead, analysts contacted trade and professional associations, employers, or industry representatives for information. On the basis of information obtained from these sources, job analysis schedules were completed in the usual manner. DEFINITION WRITING FOR THE DOT The job analysis schedules produced from 1965 to 1976, intended for use in compiling the fourth edition, were filed in the North Carolina field center by third edition DOT code, along with all the other materials resulting from establishment studies. Definition writing was not an ongoing process, and fourth edition definitions were written during 1976, the year preceding publication. (The Definition Writer's Manual, a technical manual to assist in this process, was issued a little more than a year earlier, in November 1974.) The national office assigned definition writing to the field centers by occupational categories (see Table 6-3~. As is readily apparent, the North Carolina field center wrote the major portion of the DOT: 74 percent of the base title definitions.
142 lo m o m o o o m m o On CC o , , , 1 o V, ' ' _ ,, ~ = ~ E c ._ .e 8 :~a sat 4 - 3 Go
Procedures Used to Produce the Fourth Edition DOT 143 In order to write definitions, analysts were provided with all the source materials available for the occupations in their assigned categories, which consisted primarily of job analysis schedules. For some titles there were also occupational code requests (job descriptions submitted to the Employment Service for coding to the DOT); copies of third edition definitions; and materials from trade associations, employers, or unions. An analysis of fourth edition source data, discussed in detail in chapter 7, reveals that the material available for each occupation varied greatly in both quantity (from zero to hundreds of job analysis schedules) and quality (from complete job descriptions to the notation "same as third editions. Analysts wrote definitions according to procedures contained in the Definition Writer's Manual (U.S. Department of Labor (1974), hereafter referred to as the Manua0. Definition writing involved four basic steps, each of which is discussed below: (1) evaluation of source data, (2) identification of related jobs for possible combination to form an occupation, (3) classification and rating of the occupation, and (4) definition writing. In step 1, if more than two or three source documents were available, analysts usually made up a worksheet (see Figure 6-8) to organize and array the data for easy inspection. At this point, the Manual (p. 6) advises analysts to "correct ratings that are clearly in error." The meaning and basis for assessing "error" are not spelled out, but, apparently, analysts were permitted to use their judgment to ensure that the job descriptions in the source data supported the ratings assigned. Since guidelines for determining error were not provided, the analyst writing the composite definition appears to have had a great deal of discretion in assessing and overriding the field work of other analysts. In step 2, source data were combined. The primary basis for doing so was "common work objective, work field, similarity of tasks, and the level of skill and responsibility involved (worker functions)" (Manual, p. 6~. Jobs with the same first three digits of the preliminary DOT code and the same work fields and/or generic titles were likely candidates for combination. Analysts were also advised to check jobs in parallel occupational divisions or groups to locate jobs for possible combination, for example, jobs in division 56 (processing of wood and wood products) with jobs in division 66 (wood machining). Although combined jobs did not have to be identical, they were supposed to be similar. With regard to worker functions in relation to data, people, and things, for instance, the Manual (p. 7) offers the following guidelines:
144 WORK, JOBS, AND OCCUPATIONS Any decision to combine jobs of three points difference or more should be reviewed carefully, for if the ratings assigned to the source data are correct, a range of as much as three points difference for a significant worker function would likely mean that the source data are not sufficiently similar to warrant combination. Similarly, with regard to GED, SVP, and other worker traits the Manual emphasizes that jobs should be combined only if their ratings are very similar. Instructions about GED (p. 8), for example, advise that Usually the GED level (of combined jobs) should be the same. However, because of differences between raters, the strong influence of employer requirements (which may vary from place-to-place and with supply of labor), and problems of rating borderline situations, a difference of one level but rarely two may be considered for combination when the other factors (especially svP and aptitudes) support the . . c .eclslon. Thus in deciding to aggregate jobs to form occupations, analysts used no single criterion. Analysts made the initial determination on the basis of the jobs' preliminary DOT codes and titles and then took account, within fairly narrow guidelines, of worker traits and the particular relationships among them. Having decided how data about individual jobs would be combined, in step 3 analysts assigned the occupation its DOT code and rated it for worker traits according to instructions in the Handbook. Because "erroneous" ratings had already been corrected and because jobs were aggregated on the basis of the similarity of their worker trait ratings, there should have been little within-occupation variation on each of these traits. Presumably, then, the rating was straightforward. Analysts appear to have eyeballed the raw data and chosen the most frequently occurring level for each trait (i.e., the modal value) rather than to have calculated other measures of central tendency (mean or median). As our analysis of the reliability of ratings in the next chapter shows, however, there appears to be rather more variability among job descriptions combined to form occupations than would be expected from this description of procedures. In step 4, which involved writing the actual definition, analysts drew most heavily on information contained in the job summary and description of tasks on the job analysis schedule. If a third edition title could be used, it was, as was its definition, with modification as required. Definitions were written according to the structured format described in chapter 2. In defining an occupation, analysts described its important tasks in detail, with particular emphasis on "responsibilities and requirements imposed upon the worker" (Manual, p. 38~. Significant worker functions were also supposed to be reflected in the definition, but worker traits did not have to be referred to directly unless
Procedures Used to produce the Fourth Edition DOT 145 they were especially pertinent to job performance. The Manual (p. 56) does counsel, however, that a "relationship must be maintained between the trait rating and the definition. For example, if temperament factor J (JUDGMENT) was rated, some indication should appear in the definition that describes the nature of the judgment required." Definitions were sent to the North Carolina field center, where they were reviewed by teams of analysts selected by the national office from each of the field centers. Review teams worked for a week at a time and attempted to review as many definitions as were available. Definitions written by one field center were reviewed by analysts from another. Aside from this restriction, analysts chose which occupations they would review and were not required to be particularly knowledgeable about them. Reviewers were permitted to make changes in the definitions they reviewed; changes from the third edition definition, however, had to be specially justified on an occupational definition transmittal form (used to record the final definition approved for the fourth edition). According to analysts involved in definition writing, reviewers were quite conservative and resistant to incorporating changes in existing third edition definitions. This claim is borne out by the Booz, Allen & Hamilton study. According to results from a random sample of 307 DOT base title occupations, 81 percent of fourth edition definitions were identical to those in the third edition. CONCLUSION Results from our field center site visits, coupled with inspection of the technical manuals used to assist analysts in the production of the DOT, lead us to certain general impressions about this process. First, instructions about how to study jobs appear to have been insufficient and inadequate. Major steps in the job analysis process did not have sufficient guidance (e.g., establishment selection). Furthermore, the manual of basic data collection procedures, the Handbook for Analyzing Jobs (U.S. Department of Labor, 1972), was not published until midway through production of the fourth edition. In the absence of clear, explicit instructions, decisions about major aspects of data collection were left entirely up to individual field centers and analysts. Although there appears to have been surprising uniformity among field centers in the way they conducted job analysis studies and fulfilled their industry assignments, there was certainly room for considerable variation, which may have adversely affected the quality and comparability of the data collected. Second, procedures used to produce the DOT were insufficiently documented. For virtually every step in the production process, there was
146 WORK, JOBS, AND OCCUPATIONS little or no information available about the criteria used in decision making or how a particular decision had been reached. Not only did this lack of documentation make it difficult to determine what was actually done, but also it means that nothing is Browner easily found out about the sampling properties, quality, and characteristics of the resulting data. The indeterminate nature of the data potentially compromises the usefulness of the DOT. Third, production of the DOT was poorly planned and coordinated. The modified procedures implemented prior to publication reportedly resulted in job analyses of lower quality than those produced during the period when standard procedures were in effect. (This conclusion is discussed in chapter 7.) Definitions were written especially hurriedly, with the likely result that source data were not fully explored or perhaps evenly and consistently aggregated or combined. Fourth, considering the procedures followed, there is some question as to whether, in developing material for the fourth edition, adequate coverage was obtained of newly emerging industries and occupations. The previous edition served, in erect, as the sampling frame for the fourth edition. Industries were assigned by using the third edition industry designations, and a major portion of the total data collection effort was spent trying to verify or update third edition occupations. These practices were efficient in some ways, but they were also rather conservative, minimizing the probability of incorporating newly emerging jobs in the DOT or of picking up changes in existing jobs. They may also have perpetuated the flaws, if any, of the previous edition. Furthermore, by relying on the third edition, little effort seems to have been made to explore the possibility of developing new or better sampling strategies. Data are available on industries and establishments at the local and national levels, for example, that might have been exploited to assign industries to field centers more systematically and to aid in the selection of establishments at the regional level. Fifth, no attempt was made to develop new job analysis methodology or to adapt existing methods to deal with a changing occupational structure. Although the methodology used provides a standardized and relatively objective means of obtaining job data, it is time consuming and not suitable for all jobs. In particular, it can be applied most practically to manufacturing jobs or, more generally, to any type of structured job that can be broken down into discrete tasks and performed over a limited amount of time. It is less suited to unstructured jobs, such as certain service jobs that entail widely varying tasks. The lack of a well-developed method for analyzing jobs that are not amenable to the usual procedures may have impaired the DOT'S coverage of these jobs as well as the
Procedures Used to Produce the Fourth Edition DOT 147 comparability of the resulting data for different types of jobs, especially structured versus unstructured ones. Existing alternative job analysis methodologies, such as task inventories and structured job analysis, should be explored in an effort to improve the consistency of job analysis and to delineate the boundaries of occupations more precisely. See McCormick (1979) and Prien and Ronan (1971) for recent summaries and reviews of these approaches.) These points raise a number of questions about the quality and characteristics of the data on which the DOT is based, in particular, the following: Were industries and jobs adequately covered? Are the data valid and reliable? To what extent did variations in procedures across field centers and analysts introduce bias and lack of comparability into the resulting data? What is the distribution of the occupational characteristics data (worker functions and worker traits)? What are the interrelationships among them? These questions and others are systematically addressed, to the extent possible given the data available to do so, in chapter 7. The California field center produced a collection of task analysis inventories in 1973. This collection was designed to be used as an aid in job analysis, to substitute an abbreviated data collection process for the standard procedure, and to provide guidelines for those not trained in job analysis.