Experimenting with the Model
The model reasonably approximates the
performance of recruiters. Below is
summarized the behavior of the system during a 221day working year.
The horizontal scale presents working days.
Two variables are graphed on the vertical scale; the number of persons in
the Applicant Pool is graphed as variable 1; and the number of persons signing
up as Recruits is graphed as variable 2. During
the oneyear simulation, the number of applicants in the Applicant Pool varies
from 0 to 4, and the recruiter is able to sign up 13 new recruits during the
year. These figures reasonably
described the recruiting process at the time.
The model worked. From
the data at hand, the model accurately reflected the work behavior of a
recruiter. It predicted the number
of applicants he prospected, and predicted the number of recruits he enlisted
into the Army. Still, it was
necessary to develop a feel for how accurately the model reflected recruiter
behavior. We focused on those
components of the model that are under the control of the recruiter to examine
how well our model described the actual system. A brief review of the basic model suggests that, while there
are relatively few such components, they might include the following:
•
Prospecting Effort, or the number of leads worked each day,
•
Sales Time, or the average number of days required to sell one prospect,
and
•
Enlistment Processing Time, or the average number of days required for
MEPS processing.
