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 41
4
Modeling
Need for and Use of Models
When attempting to understand interactions within a complex sys-
tem, it is frequently useful to construct one or more simplified repre-
sentations of the system under study. These models may address only
one aspect of the problem, or they may reflect only its structure.
Models of this type, although limited in capability, are useful because
they offer controlled "what if" capability, insight into critical relation-
ships, evaluation of assumptions, and a low-cost study environment.
In creating a model, many simplifying assumptions must lee made,
and the available data must be critiqued, compared, and consolidated.
These two interrelated activities focus the attention of the investiga-
tors on the more fundamental aspects of the problem being studied.
Certainly they force attention toward the measurable attributes of the
problem. The adequacy of available data reflects the broad consensus
viewpoint and the definition of historically significant variables.
Critique of Existing Models
A large number of models and studies have been made of various
aspects of engineering manpower. These may lie classified in two gen-
eral categories supply and demand. On the supply side, for example,
it has been shown that disciplines chosen lay students are influenced lay
future job outlooks. On the demand side, future jolt openings are related
41
OCR for page 42
42
INFRASTRUCTURE DIAG~MMING AND MODELING
to the pattern of economic growth. These are but two examples chosen
from literally hundreds of papers and books.
More recent studies employ more elegant methods. In no small part
this reflects the availability of better-quality data from a diverse set of
sources including the National Science Foundation {NSF, the Bureau
of Labor Statistics {BLS), and the Engineering Manpower Commission
{EMC~. Specific models are described in the proceedings of the Sympo-
sium on Labor-Market Conditions for Engineers sponsored by the
National Research Council. ~
Although many of the prior studies improve our understanding of
what has occurred relative to the flow of graduates from the education
arena to the workplace, we know painfully little about the "Why's"
behind this movement Why is a curriculum chosen? Why is one job
accepted while another goes begging? Why do people move between the
government, business, and education sectors? And so on. An additional
complication in modeling is that it takes three or more years for the
supply of newly graduated engineers to adjust to changes in demand.
This can also be complicated by cyclical overcorrections.
Developing the CEUE Simulation Model
The Panel on Infrastructure Diagramming and Modeling found that
existing models did not describe the engineering community in suffi-
cient detail to analyze the flows described in the flow diagram dis-
cussed in Chapter 3. The panel anticipates that the NSF model will
serve that purpose in the long term. However, the panel decided to
develop an interim simulation model referred to here as the CLUE*
model-as an aid in analyzing flows for the purpose of this study.
Model Objectives
Avery restricted but still interesting and challenging set of objectives
was chosen for the CEUE model described here. The pane! first concen-
trated on the flow of the population-to-education-to-job-market sup-
ply. {This emphasis was chosen because it was consistent with the flow
diagram responsibility of the panel. ~ Second, the model was limited to a
relatively high summary level because of the limitations of the avail-
al~le data and in order to ease the calibration and maintenance prob-
lems. Thus, the CEUE model does not give detail on engineering
disciplines. Third, it was decided to study only the flows of people who
* CEUE = Committee on the Education and Utilization of the Engineer.
OCR for page 43
MODELING
43
have received their formal education in engineering and scientific dis-
ciplines. Finally, the model runs in an open loop mode, which allows
easier interaction and model formulation but obscures unsuspected
feedback phenomena. The panel suspects that an upper bound on engi-
neering enrollment will occur during the early 1 990s because of a short-
fall in new Ph.D.s choosing to teach in engineering. This remains only
a supposition at this time, because the number of students was not
limited by the number of anticipated trained and working teachers.
Thus, it was assumed that there would be a significant improvement in
teaching productivity over the next 20 years. It is recognized that the
model is based on the size and flows among the several populations and
does not directly consider economic and political forces that may influ-
ence allocation of resources and thus size and flows among the various
population pools. However, dependable data are not available to
directly relate the size and flows of population pools in the engineering
community to economic factors. Rather, the size of pools and flows
themselves represents the effects of economic, political, and social
factors.
In summary, the CLUE model can simulate the flow of engineers in
the United States beginning in 1950. It is possible to run alternative
cases, thus giving relatively crude forecasts of the supply of engineers to
the work force.
Program Features
The CEUE simulation models was written to run on the IBM Per-
sonal Computer PC. Basically, the program has two main parts: his-
torical and forecasting. The historical module uses historical data
obtained from many sources and produces on the color monitor graphic
representations of these data. There are also various options available
for the user that permit trend analysis. These options are invoked from
a user-friendly menu. For example, pressing a function key on the PC
keyboard results in the production of a graph on the color monitor.
Also, with a graphic printer, a user can obtain a hard copy of the result.
The forecasting module predicts the results listed below based on
population distribution and growth. The program starts with an initial
formulation of the population data, which are obtained from various
historical sources. Then it forecasts the following:
· the population distribution {this includes males and females,
ranging in age from 1 to 100~;
· high school graduates;
OCR for page 44
44
INFRASTRUCTURE DIAGRAMMING AND MODELING
.
first-year enrollment in colleges;
· l~achelor's degrees in engineering, physical science, mathematics,
and computer science;
· M.S. graduates in the aforementioned fields;
· Ph.D. graduates in the aforementioned fields {the graduates are
divided into classes: males, females, and U.S. citizens and foreign
nationals); and
· the engineering work force, in different fields, and by different
employers.
The forecasting module is also menu-driven and, like the historical
module, uses function keys to choose options. The forecasting module
first computes the results and then asks the user whether he or she
wants to view printed answers on the monochrome monitor or plotted
answers on the color graphics monitor. For example, a graph can lie
displayed that shows the decline in the number of the high school
graduates starting in the early 1980s and extending through the mid-
1990s. This downward trend affects the number of B.S., M.S., and
Ph.D. graduates in the different fields. The time delays between B.S.,
M.S., and Ph.D. degrees are evident from the displayed graph.
Finally, the model has a "what if" option. For example, a hypotheti-
cal case can lie presented in which the graduation rates of females and
foreign nationals are used as parameters to analyze the impact their
potential increased enrollment might have on the declining supply of
. . .
engineering grac .uates.
Results, Self-Critique, and Likely Extensions
The total U.S. population lay age and sex provides the basis for esti-
mating the number of individuals reaching college entry age. Projected
values of the key variables used in this simulation model are generally
leased on the naive assumption of no future changes in their value.
However, the population projections reflect the declining birth rates
that were experienced beginning in the late 1950s and continuing to the
present time.3
Based on these population projections, a steady decline in the future
annual 18-year-old cohorts to about 75 percent of current numbers
could result in a decline in the number of future engineering graduates.
This decline could be offset lay factors associated with increases in the
number of students choosing engineering. For example, historical data
indicate that 4 to 6 percent of male high school graduates have pursued
engineering degrees. The panel estimates that an increase in the rate
OCR for page 45
MODELING
45
approaching the higher range could offset the overall decline in student
population.
The CLUE model offers an opportunity to evaluate data on women
and to assess the implications of their growing participation in engi-
neering [see Figure 7J. The rapidly increasing female participation in
computer-related disciplines is a likely harbinger of similar trends in
engineering. A number of the more mathematically leased engineering
disciplines already indicate a meaningful increase in female participa-
tion. Overall, the percentage of female engineering graduates increased
lay a factor of 10 during the 1970s. Changes in minority participation
can be sized lay changing overall participation rates. However, the his-
torical data have not been included at this time.
The model can also analyze the effects of the increasing percentage of
engineering doctoral graduates who are foreign-born. Today these
foreign-l~orn graduates are approaching 50 percent of the number of
Ph.D.s produced by universities. It is interesting to note that this effect
is largely the result of a significant decrease in the number of U. S.-born
students continuing beyond the master's level.
The model can also evaluate the replacement rate for engineering
Ph.D.s working in the education sector. Even if the current supply of
new engineering faculty were adequate, our universities are expected
loo
90
80
70
N
G 60
50
40
30
20
10 _
O 1 1 1 1 1 1 ~1 1 1
1970 1975 1980
EXPECTED SUPPLY ~ FEMALE
IMPROVEMENT TO DOUBLE
THEIR 1984 ROTE
ACTUAL
~ /
PROJECTED SUPPLY DUE TO
EXPECTED DEMOGRAPHICS
~1 1 1 1 1 1
1985
Y EARS
1 1 1 1 1 1 1 1 1 1 1 1
1990 1995 2000
FIGURE 7 B.S. engineering supply (CLUE Model).
OCR for page 46
46
INFRASTRUCTURE DIAGMMMING AND MODELING
to lace an increasing demand for new faculty to replace the growing
number of current faculty who will lee retiring during the next 15 years.
On a discipline-l~y-discipline basis, this mismatch may be even more
serious.
The model does not produce projections with narrow confidence
lands. For example, using the actual 1950 statistics as of 1950 and
projecting for 30 years gives actual-to-projection ratios of 50 to 100
percent. Almost all variables changed lay 25 percent or more over the
period, and many moved lay a factor of 2. Thus, exercising the CLUE
model with the naive assumptions described earlier would lie inade-
quate for many purposes regardless of how refined or detailed the model
was. On the other hand, isolation of the variable for which model
results are particularly sensitive enables the model to offer construc-
tive direction to those guiding future causal studies and to those seek-
ing new education policies.
Notes
1. Office of Scientific and Engineering Personnel, National Research Council. Labor-
Market Conditions for Engineers: Is There a Shortage: Proceedings of a Sympo-
sium. (Washington, D. C .: National Academy Press, 19841.
2. Detailed data on the model and its structure and sample runs are available on
request from the Office of Scientific and Engineering Personnel, National Research
Council, 2101 Constitution Ave., N.W., Washington, D.C. 20418.
Note, however, that even though the methodology is sound, birth rate patterns
have shifted drastically over the past three decades, and the model will merely
reflect our current estimate of future birth rates. For present purposes this is of
little importance, since all new entries into the engineering work force during this
century have already been born.
Representative terms from entire chapter:
clue model
