This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
A work standard is the time required for a trained worker to perform a task following a prescribed method with normal effort and skill
Used in the following ways: Establishing prices and costs Motivating workers Comparing alternative process designs Scheduling Capacity planning Performance appraisal
Time study is the method used most often Step 1: Selecting work elements
Step 2: Timing the elements
Step 3: Determining sample size2
tpz
n
wheren =required sample sizep =precision of the estimate as a proportion of the true valuet =select time for a work element =standard deviation of representative observed times for a work elementz =number of normal standard deviations needed for the desired confidence
Estimating the Sample Size in a Estimating the Sample Size in a Time StudyTime Study
EXAMPLE H.1
A coffee cup packaging operation has four work elements. A preliminary study provided the following results:
Work Element
Standard Deviation, , (min)
Select Time, t, (min)
Sample Size
1. Get two cartons 0.0305 0.50 5
2. Put liner in carton 0.0171 0.11 10
3. Place cups in carton 0.0226 0.71 10
4. Seal carton and set aside 0.0241 1.10 10
Work element 1 was observed only five times because it occurs once every two work cycles. The study covered the packaging of 10 cartons. Determine the appropriate sample size if the estimate for the select time for any work element is to be within 4 percent of the true mean 95 percent of the time.
Determining the Normal TimeDetermining the Normal Time
EXAMPLE H.2
Suppose that 48 additional observations of the coffee cup packaging operation were taken and the following data were recorded:
Work Element t F RF
1 0.53 0.50 1.05
2 0.10 1.00 0.95
3 0.75 1.00 1.10
4 1.08 1.00 0.90
Because element 1 occurs only every other cycle, its average time per cycle must be half its average observed time. That is why F1 = 0.50 for that element. All others occur every cycle. What are the normal times for each work element and for the complete cycle?
Management needs a standard time for the coffee cup packaging operation. Suppose that A = 0.15 of the normal time. What is the standard time for the coffee cup packaging operation, and how many cartons can be expected per 8-hour day?
Lucy and Ethel have repetitive jobs at the candy factory. Management desires to establish a time standard for this work for which they can be 95% confident to be within ± 6% of the true mean. There are three work elements involved:
SOLUTION
Step 1: Selecting work elements
#1: Pick up wrapper paper and wrap one piece of candy
#2: Put candy in a box, one at a time
#3: When the box is full (4 pieces), close it and place on conveyor
Step 2: Timing the elements. Select an average trained worker, Lucy will suffice.
Element
Initial Observation Cycle Number, Minutes
1 2 3 4 5 6 7 8 9
Wrap #1: .10 .08 .08 .12 .10 .10 .12 .09 .11
Pack #2: .10 .08 .08 .11 .06 .98* .17 .11 .09
Close #3: .27 ... ... ... .34 ... ... ... .29
SelectTime, t
Standard Dev,
0.1 0.015
0.1 0.03295
0.3 0.03606
Step 3: Determining sample size. First calculate t for each element in Step 2. Assume a 95% confidence interval, with z = 1.96. The precision interval of ± 6% of the true mean implies p = 0.06. To determine the sample size, use the largest value of / t .
Element Select Time, t Frequency Rating Factor Normal Time
Wrap #1: 1.00 1.2
Pack #2: 1.00 0.9
Close #3: 0.25 0.8
Application H.1Application H.1
d. Subjectively determine the proportion of the normal time to be added for allowance, and then calculate standard time ST. Let the allowance be 18.5% of the normal time (A = .185).
Most frequently used method for setting time standards
Qualified analysts can typically set reasonable standards Not appropriate for “thinking” jobs Not appropriate for non-repetitive jobs Inexperienced persons should not conduct time
studies because errors can result in unreasonable standards
Workers may object to judgment and subjectivity involved
Elemental Standard Data ApproachElemental Standard Data Approach
Useful for processes with high divergence, but when a high degree of similarity exists for basic elements of work for different services and processes Time standards are developed for common work
elements Study results are stored in a database for later use in
establishing standards for jobs requiring those elements Allowances must still be added An equation may be used to account for the effect on
time required by certain variable characteristics of the jobs
This approach reduces the number of time studies needed, but does not eliminate time studies
Select a sample size so that the estimate of the proportion of time spent on a particular activity that does not differ from the true proportion by more than a specified error, so
eppep ˆˆˆ
where=sample proportion (number of occurences divided by the sample size)e =maximum error in the estimate
The hospital administrator at a private hospital is considering a proposal for installing an automated medical records storage and retrieval system. To determine the advisability of purchasing such a system, the administrator needs to know the proportion of time that registered nurses (RNs) and licensed vocational nurses (LVNs) spend accessing records. Currently, these nurses must either retrieve the records manually or have them copied and sent to their wards. A typical ward, staffed by eight RNs and four LVNs, is selected for the study.
a. The hospital administrator estimates that accessing records takes about 20 percent of the RNs’ time and about 5 percent of the LVNs’ time. The administrator wants 95 percent confidence that the estimate for each category of nurses falls within 0.03 of the true proportion. What should the sample size be?
b. The hospital administrator estimates that the annual amortization cost and expenses for maintaining the new automated medical records storage and retrieval system will be $150,000. The supplier of the new system estimates that the system will reduce the amount of time the nurses spend accessing records by 25 percent. The total annual salary expense for RNs in the hospital is $3,628,000, and for LVNs it is $2,375,000. The hospital administrator assumes that nurses could productively use any time saved by the new system. The pilot work sampling study resulted in the data shown in Figure H.2. Should the administrator purchase the new system?
a. Using estimates for the proportion of time spent accessing records of 0.20 for RNs and 0.05 for LVNs, an error of ± 0.03 for each, and a 95 percent confidence interval (z = 1.96), we recommend the following sample sizes:
Eight RNs and four LVNs can be observed on each trip. Therefore, 683/8 = 86 (rounded up) trips are needed for the observations of RNs, and only 203/4 = 51 (rounded up) trips are needed for the LVNs. Thus, 86 trips through the ward will be sufficient for observing both nurse groups. This number of trips will generate 688 observations of RNs and 344 observations of LVNs. It will provide many more observations than are needed for the LVNs, but the added observations may as well be recorded as the observer will be going through the ward anyway.
b. Before using the estimates from the work sampling study, we must be sure that additional sampling is not required. Figure H.2 shows that RNs accessed records 124 times and LVNs only 28 times. The computer output shows that the proportion of working time spent on accessing records is 0.1802 for the RNs and 0.0814 for the LVNs. Thus, the original estimates were off the mark. The computer uses the new estimates for the proportions in the same formulas we used in part (a) to revise the sample sizes. However, the new sample sizes are smaller than those already used, so no additional sampling is required. If the sample sizes were too small for the proportions found, additional sampling would have to be performed. In addition, the confidence interval shows the range possible in the “true” proportions, based on the results of the pilot study. For example, the actual proportion of time spent by the RNs on accessing records could be as low as 0.15 and as high as 0.21.
Because the nurses will not be using the system all the time, we accept the supplier’s estimate of 25 percent to determine the value of the time spent accessing records. Estimated annual net savings from the purchase of the automatic medical records storage and retrieval system are
Net savings = 0.25[($3,628,000)(0.18) + ($2,375,000)(0.08)] - $150,000
Major League Baseball (MLB) is concerned about excessive game duration. Batters now spend a lot of time between pitches when they leave the box to check signals with coaches, and then go through a lengthy routine including stretching and a variety of other actions. Pitching routines are similarly elaborate. In order to speed up the game, it has been proposed to prohibit batters from leaving the box and to prohibit pitchers from leaving the mound after called balls and strikes. MLB estimates the proportion of time spent in these delays to be 20% of the total game time. Before they institute a rules change, MLB would like to be 95% confident that the result of a study will show a proportion of time wasted that is accurate within ± 4% of the true proportion.
Advantages No special training required of observers Several studies can be conducted simultaneously More economical for jobs having long cycle times Workers prefer this method to time studies
Disadvantages A large number of observations are required Usually not used for repetitive, well-defined jobs Workers may increase quantity at the expense of quality
Managers should carefully evaluate work measurement techniques to ensure that they are used in ways that are consistent with the firm’s competitive priorities
Technological changes Increased automation There is less need to observe and rate worker
performance, because work is machine paced Work sampling may be electronically
For a time study of a health insurance claims-adjusting process, the analyst uses the continuous method of recording times. The job is divided into four work elements. Shown in Figure H.3 are the performance rating factors, RF, and the continuous method recorded times, r, for each work element.
a. Calculate the normal time for this job.
b. Calculate the standard time for this job, assuming that the allowance is 20 percent of the normal time.
c. What is the appropriate sample size for estimating the time for element 2 within ± 10 percent of the true mean with 95 percent confidence?
a. To get the normal time for this job, we must first determine the observed time, t, for each work element for each cycle. We calculate the time for each observation by finding the difference between successive recorded times, r. For example, the time for the fifth observation of the first work element is the difference between the recorded time when that element was completed and the time when the fourth observation of the fourth work element was completed. With no extreme variation in the observed times for the work elements, they are representative of the process. All the data can be used for calculating the average observed time, called the select time, t, and the standard deviation of the observed times, . The results of those calculations are given in Figure H.3. Every work element occurs during every cycle, so the frequency, F, equals 1.
A library administrator wants to determine the proportion of time the circulation clerk is idle. The following information was gathered randomly by using work sampling:
DayNumber of Time
Clerk BusyNumber of Time
Clerk IdleTotal Number of
Observations
Monday 8 2 10
Tuesday 7 1 8
Wednesday 9 3 12
Thursday 7 3 10
Friday 8 2 10
Saturday 6 4 10
If the administrator wants a 95 percent confidence level and a degree of precision of ± 4 percent, how many more observations are needed?
The total number of observations made was 60. The clerk was observed to be idle 15 times. The initial estimate of the sample proportion is . The required sample size for a precision of 4 percent is
2
2 1e
ppzn
ˆˆ 2
2
040
750250961
.
...
nsobservatio 451 or 450.19,
As 60 observations have already been made, an additional 391 are needed.