WORK SAMPLING Lecturer: S.U.Seçkiner Source: Motion and Time Study Design and Measurement of Work

Key points Work Sampling is a method for analyzing work by taking a large number of observations at random times. Use work sampling to: –Determine machine utilization –Determine allowances –Establish time standards Use as many observations as practical Take observations at random times over two or more weeks.

Work sampling is a technique used to investigate the proportions of total time devoted to the various activities that constitute a job or work situation. The results of work sampling are effective for determining machine and personnel utilization, allowances applicable to the job and production standards. Work sampling frequently provides the same information faster and considerably less cost than time study.

In conducting work sampling studies; Large number of observations at random intervals, The ratio of observations of a given activity to the total observations approximates the percentage of time that the process is in that state of activity.

Work sampling was first applied in the British textile industry Later, technique was brought to the United States. The accuracy of the data determined by work sampling depends on the number of observations and the period over which the random observations are taken

Advantages 1.It does not require continuous observation by an analyst over a long period of time 2.Clerical time is diminished, 3.The total work-hours expended by the analyst are usually much fewer, 4.The operator is not subjected to long period stopwatch observations. 5.Crew operations can be readily studied by a single analyst.

Work sampling has three main uses; 1. Activity and delay sampling (to measure the activities and delays of workers or machines. Ex:percentage of working 2. Perfomance sampling (to measure working time and nonworking time of a person on a manual task, and to establish a performance index or performance level for the person during his or her working time. 3. Work measurement (under certain circumstances, to measure a manual task, that is, to establish a time standard for an operation.

Simple example of work sampling The work sampling procedure in its simplest form consists of making observations at random intervals of one or more operators or machines and noting whether they are working or idle. For example; –Working=36 –Idle=4 –Total=40 observations –Percentage of idle time=(4/40)*100=10% –Percentage of working time=(36/40)*100=90% –If one operator performed task for an 8 hour day. Operator was idle 48 minutes of the day (480*0,10=48) and was working 432 minutes of the day (480*0,90=432).

The Theory of Work Sampling The theory of work sampling is based on the fundamental law of probability: at a given instant, an event can be either present or absent. Random sampling requires that there be no bias in the sampling process. It is important that the concept of randomness be understood and carefully followed in work sampling studies.

The normal distribution curve The normal distribution curve is typical of the kind of frequency distribution which is of importance in work sampling because it represents graphically the probabilty of the occurence of certain chance phenomena.

Confidence level At the outset it is necessary to decide what level of confidence is desired in the final work sampling results. The most common interval is 95% The area under the curve at 2 sigma or two standard deviations is %. This means that the probability is that 95% of the time the random observations will represent the facts and 5 percent of the time they will not. The formula for determining the sample size for a confidence level of 68 percent or 1 sigma is S=desired relative accuracy, p=percentage expressed as a decimal, N=number of random observations (sample size)

Accuracy of work sampling measurement For many kinds of measurement an accuracy of +,-5 percent is considered satisfactory. This is sometimes referred to as the standard error of the percentage. Following illustrations we will assume that a confidence level of 95% and an accuracy of +,-5% satisfactory. Also assuming that the binomial distribution is used as the basis for determining the error, the formula for determining the number of observations required is S=desired relative accuracy p= percentage occurence of an activity or delay being measured, expressed as a percentage of the total number of observations or as a decimal N=Total number of observations (sample size)

Example Suppose we want to determine the percentage of idle time of the automatic screw machines in a department. CL=95% error=+,-5 % We want to know how many random observations will be needed to give us the desired results. 100 observations were made and preliminary study 25 observations showed the machines to be idle. (25%)

Solution N=? p=25% S=+,-5%=+,-0.05 =

Fig 245

Determination of Accuracy Desired accuracy must be lower than 5%. For example, if we calculates S is 3.5% then it is sufficient result. The accuracy or standard error of +,-3.5 percent means that the results is correct within +,-3.5 percent of ex:30 percent (+,-3.5%*30%=1.05%) or the value was between 28.95% and percent. The 95% CL means that the probability is that in 95 cases of 100 the above results will represent the facts.

Determining Time Standards by Work Sampling Sampling can be used for measuring work as well as for measuring delays, idle times and performance. Sampling can be used instead of time study for long activities. It is possible to determine by work sampling the percentage of the day that a person is idle and the percentage of the day that he is working, as well as the average performance index or speed at which he worked the working portion of the day. For example, 8-hour day as a drill-press operator. A work sampling study might show 15% idle or 72 minutes (480*0.15=72) and 408 minutes (480*0.85) At an average performance index of 110 percent. During the day 420 pieces

Establishing Time Standards by Work Sampling

Some Advantages and Disadvantages of Work Sampling in Comparison with Time Study Many operations or activities which are impractical or costly to measure by time study can readily be measured by work sampling. A simultaneous work sampling study of several operators or machines may be made by asingle observer. Ordinarily an analyst is needed for each operator or machine when continous time studies are made. It usually requires fewer man-hours and costs less to make a work sampling study than it does to make a continous time study.

Some Advantages and Disadvantages of Work Sampling in Comparison with Time Study Ordinarily work sampling is not economical for studying a single operators or machines located over wide areas. Time study permits a finer breakdown of activities and delays than is possible with work sampling. Work sampling cannot provide as much detailed information as one can get from time study. Management and workers may not understand stastical work sampling as readily as they do time study.