WORK SAMPLING
1
Assume that in your organization material is moved
with a fork truck. There seem to be long delays.
Upon what do you base your opinion? The question
is asked, “Do we need another truck or the present
one idle too much?” This lead to a more specific
question: “What is the present utilization of the
truck?”
The purpose of work sampling is to obtain
information in order to make decisions.
2
We might have someone to follow the truck for a
specific time period and record the following type of
information for each day:
– 0700 Went to shipping dock
– 0706 Parked outside supervisor’s office
– 0711 Left office with orders
– 0712 Entered first freight car with load
This would be a continuous time study giving a
complete picture of the situation. The problem is the
3 expense of the study.
Therefore, discrete sampling could be used instead
of continuous sampling.
Work sampling permits gaps between sample
readings. As the name suggests, work sampling
involves a portion or sample of work activity. Then,
based on the findings in this sample, statements can
be made about the activity.
Cheaper and easier approach to work measurement
4
For example, if we were to observe our fork truck at
100 random times during the day and it was involved
in a transportation mission for 30 of the 100 times,
we would estimate that our fork truck spends 30% of
its time directly on transportation activities.
However, observing an activity even 100 times may
not provide the accuracy desired in the estimate.
5
Required Number of Observations
Optimal number of observations depends on:
– Desired absolute accuracy
– Proportion of occurrence
– Confidence level desired
6
Types of Accuracy
Relative: Percentage of the proportion
Absolute: Measurement of the proportion
If proportion is .4, 10% relative accuracy =
4% absolute accuracy.
7
Confidence Level
Confidence expresses the long-run
probability that the sample mean is within the
accuracy limits.
8
The formula to determine the number of
observations required is:
2
z
n= p(1 - p)
e
where
n = sample size (number of sample observations)
z = number of standard deviations from mean for desired
level of confidence
e = degree of allowable error in sample estimate
p = proportion of time spent on a work activity estimated
9 prior to calculating work sample
Example
Estimated idle percent of fork truck = 40%
Desired relative accuracy = ±10%
Desired confidence level = 68%
Solution: .1 × .4 = 4%=e
68% confidence = ±1σ z=1
Solving equation, n = 150
p=.4
Statement: If the sample average of 150 observations
is 40%, you can say with 68% confidence that the
long-run idle percentage is between 36% and 44%
with 40% being the most likely estimate, if the
10 situation does not change.
The Three Primary Applications for
Work Sampling
1. Ratio delay to determine the activity-time
percentage for personnel or equipment: For
example, management may be interested in the
amount of time a machine is running or idle.
2. Performance measurement to develop a
performance index for workers. When the amount
of work time is related to the quantity of output, a
measure of performance is developed. This is
useful for periodic performance evaluation.
3. Time standards to obtain the standard time for
task. When work sampling is used for this purpose,
however, the observer must be experienced
because he or she must attach a performance
11 rating to the observations.
Steps to Making a Work Sampling
Study
1. Identify the specific activity or activities that are the main
purpose for the study. For example, determine the
percentage of time that equipment is working, idle, or under
repair.
2. Estimate the proportion of time of the activity of interest to
the total time (e.g., that the equipment is working 80% of
the time). These estimates can be made from the analyst’s
knowledge, past data, reliable guesses from others, or a
pilot work-sampling study.
3. State the desired accuracy in the study results.
4. Determine the specific times when each observation is to
be made.
5. At two or three intervals during the study period, re-
compute the required sample size by using the data
collected thus far. Adjust the number of observations if
12 appropriate
Work Sampling Applied to Nursing
There has been a long-standing argument that a
large amount of nurses’ hospital time is spent on
non-nursing activities. This creates an apparent
shortage of well-trained nursing personnel, wastes
talent, hinders efficiency, and increases hospital
costs because nurses’ wages are the highest single
cost in the operation of a hospital. Further, pressure
is growing for hospitals and hospital administrators
to contain costs. With that in mind, let us use work
sampling to test the hypothesis that a large portion of
nurses’ time is spent on non-nursing duties.
13
Assume at the outset that we have made a list of all the
activities that are part of nursing and will make our observations
in only two categories: nursing and non-nursing activities
We estimate that nurses spend 60% of their time in nursing
activities
We would like to be 95% confident that findings of our study are
within the absolute error range of ±3% (We want to be 95%
confident that the true percentage lies between 57 and 63%
We find that 1,024 observations are required. If our study is to
take place over ten days, we start with 103 observations per
day.
To determine when each day’s observation are to be made, we
assign specific numbers to each minute and use a random
number generator to set up a schedule.
14
Time Assigned Numbers
7:00 – 7:59 A.M. 100-159
8:00 – 8:59 A.M. 200-259
9:00 – 9:59 A.M. 300-359
10:00 – 10:59 A.M. 400-459
11:00 – 11:59 A.M. 500-559
12:00 – 12:59 P.M. 600-659
1:00 – 1:59 P.M. 700-759
15 2:00 – 2:59 P.M. 800-859
Random number Corresponding Time
669 nonexistent
831 2:31 P.M.
555 11:55 A.M.
470 nonexistent
113 7:13 A.M.
080 nonexistent
520 11:20 A.M.
204 8:04 A.M.
732 1:32 P.M.
420 10:20 A.M.
16
Observation Schedule Time Nursing Activity Non-nursing Activity
1 7:13 A.M.
2 8:04 A.M.
3 10:20 A.M.
4 11:20 A.M.
5 11:55 A.M.
6 1:32 P.M.
7 2:31 P.M.
17
To be perfectly random in this study, we should also
“randomize” the nurse we observe each time.
At the end of the second day (and 206 observations), we
decide to check for the adequacy of our sample size. Let us say
we made 145 observations of nurses working and 61 of them
not working, which gives 70.4% working.
This corresponds to 890 observations. Because we already
have taken 206 observations, we need take only 684
observations over the next 8 days, or 86 per day.
When the study is half over, another check should be made.
Another computation should be made before the last day to see
if another adjustment is required.
If after the 10th day several more observations are indicated,
18 these can be made on day 11.
If at the end of the study we find that 66% of
nurses’ time is involved with what has been
defined as nursing activity, there should be
an analysis to identify the remaining 34%.
Approximately 12 to 15% is justifiable for
allowances, which leaves 19 to 22% of the
time that must be justified and compared to
what the industry considers ideal levels of
nursing activity.
19
Work Sampling Compared to Time Study
Work sampling offers several advantages:
Several work sampling studies may be conducted
simultaneously by one observer
The observer need not be a trained analyst unless
the purpose of the study is to determine a time
standard
No timing devices are required
Work of a long cycle time may be studied with fewer
observer hours
20
The duration of the study is longer, which minimizes
effects of short-period variations
The study may be temporarily delayed at any time
with little effect
Because work sampling needs only instantaneous
observations (made over a longer period), the
operator has less chance to influence the findings by
changing work method
21
Deriving a Time Standard Using Work
Sampling
Information Sources of Data Data for One Day
Total time expended by nurse (A) Computer payroll system 480 min.
Number of patients transferred (B) Planning department 42 patients
Working time in percent (C) Work sampling 80%
Idle time in percent (D) Work sampling 20%
Average rating (E) Work sampling 110%
Total allowances (F) Company time-study manual 20%
Standard time per patient = ((A x C x E) / B) x (1 / 1 – F) = ((480 x 0.80 x 1.10) / 42) x (1 / (1-.20)) =
12.57 min.
22
