should be chosen for each contract requirement shown on the EPRS, Figure
4-1, and it is not unusual to find a variety of surveillance methods in use
for the inspection of any one contract.
100 percent inspection is preferred when:
4-420 100 Percent Inspection.
(a) The total number of services is small,
(b) The satisfactory performance of a work requirement is critical,
(c) The expense of 100 percent sampling can be justified.
Examples of the proper use of 100 percent inspection are: ambulance and
police vehicle response time; and daily cleaning of key public rooms.
4-430 Planned Sampling. Planned sampling is preferred when:
(a) The total number of services is small.
(b) Specified contract locations must be monitored due to individual
importance, for example, galley garbage containers as opposed to those in
remote administrative areas. It is also useful when the contractor has
performed poorly in certain defined areas.
(c) The contract requirement does not need to be as vigorously
inspected as if 100 percent or random sampling was used.
(d) Special consideration is given to areas with command interest or
where unsatisfactory performance has been observed during the previous
evaluation period.
(e) Follow up on customer complaints and random sampling is required.
(f) Attention is focused on known problem areas, thus providing the
contractor with a greater incentive to improve performance because he knows
that he is being observed.
4-440 Random Sampling. Random sampling is preferred when:
(a) Inspection resources are limited,
(b) The population is large and relatively homogenous,
(c) Contract requirements occur frequently or continuously,
(d) There is a well documented audit trail of contractor activities
such as completed work orders or vehicle maintenance logs to evaluate.
Examples of the proper use of random sampling includethe evaluationof
repetitive tasks associated with janitorial work, building and grounds
maintenance, and guard services.
74