|
Items in cart [0] |
Questions? Call 888-624-8373 |
| Copyright © 2009. National Academy of Sciences. All rights reserved. Terms of Use and Privacy Statement |
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 114
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
OCR for page 115
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
OCR for page 116
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
OCR for page 117
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).
OCR for page 118
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
OCR for page 119
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
OCR for page 120
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.
OCR for page 121
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.
OCR for page 122
OCR for page 124
OCR for page 137
OCR for page 138
OCR for page 139
OCR for page 140
OCR for page 141
OCR for page 142
OCR for page 143
OCR for page 144
OCR for page 145
OCR for page 146
OCR for page 147
Representative terms from entire chapter:
fourth edition
122
Cq
To
P:
o
~ o
..Q ·-
o ~
~ .
~¢
c3 ~
B
Ha.
~ _
~ it,
¢
Em
-
¢
o
C)
C.
of
to
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
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.
