Pemeriksaan Total.
Sebagai teknik pengendalian mutu. pemeriksaan total melibatkan kelengkapan dan pemeriksaan total pekerjaan yang diproduksi oleh masing-masing karyawan untuk menentukan ya atau tidaknya standar mutu minimum telah dicapai. Jika bukan, ukuran mengoreksi barangkali akan diambil.
Pemeriksaan Total diinginkan untuk tertentu jenis pekerjaan ketatausahaan. Seperti contoh yang umum pemeriksaan total adalah koreksi cetakan pekerjaan diketik. Lain contoh pekerjaan ketatausahaan yang sering menerima total pemeriksaan adalah verifikasi kalkulasi seperti ilmu hitung penting dan hasil menyusun data statistik.
Oleh karena itu sifat alami beberapa bentuk pekerjaan ketatausahaan, pemeriksaan total mungkin tidak perlu. Dalam beberapa peristiwa, pekerjaan akan menjadi sangat terbatas. Di lain kejadian, sangat kecil kesempatan kesalahan yang telah dibuat. Contoh penyimpanan surat menyurat klien. Walaupun beberapa surat menyurat mungkin tidak tersimpan, situasi ini tidak menjamin keabsahan pemeriksaan total file untuk memastikan ketelitian menyimpan.
QUALITY CONTROL TECHNIQUES
Several techniques have been developed for maintaining quality control. Among these are total inspection, spot checking, statistical quality control, and Zero Defects.
Total Inspection. As a quality control technique. Total inspection involeves complete and total inspection of work produced by each employee to determine whether or not minimum quality standards have been attained. If not, corrective measures will perhaps have to be taken.
Total inspection is desirable for certain types of office work. Perhaps the most common example of total inspection is the proofreading of typewritten work. Other example of office work that frequently receive total inspection are the verification of important arithmetical calculations and the results of compiling statistical data.
Because of the nature of some types of office work, total inspection may not be necessary. In some instances, the work is of limited importance. In other instances, there is very little chance that errors have been made. An example is the filing of clients correspondence. Although some of the correspondence may be misfiled, this situation does not warrant the total inspection of the files to insure filing accuracy.
Spot checking. The spot checking technique involves periodically checking the quality of an employee’s work. The desirability of this technique is frequently challenged because the work is spot checked and no use is made of statistical processes to determine how much, who, and when the checks are to be made. By adding the statistical dimension, which is a characteristic of statistical quality control, more valid results are possible.
Statistical quality control. Because total inspection of work is not always desirable nor is spot checking sufficiently accurate, statistical quality control is used. This technique produces accurate and reliable results because of the statistical sampling base that it uses.
Certain fundamental statistical elements are found in statiscal quality control. These elements include sampling, normal distribution, and control limits. Sampling, which is based on laws of probability, is used to determine what percent of the total output is as error-free as the sample. In other words, if proper statistical procedures are used, the quality of the sample should be representative of the quality of the whole.
Statistical tables, which are available for determining appropriate sampe size, take into consideration the following elements: the total number of units from which the sample is being drawn and the minimum accuracy level that is required. If the total output consisits of one hundred units and an accuracy level of 90 precent is acceptable, fewer samples will have to be studied than when an accuracy standard of 95 precent is required. To illustrate if the total output consists of one hundred units and it is determined that 96 precent accuracy is required, the statistical tables indicate that fifteen of the units should be randomly selected for inspection. If no errors are found within the fifteen units that are also free of errors. But if one error is found withinthe sample of fifteen units, according to the law of probability, seven of the one hundred units could be expected to contain errors.
Normal distribution is another of the elements involved in statistical quality control. Normal distribution is based on the principle that randomly observed occurrencesof of a sufficient quantity tend to be distributed around the mean or average of all the occurrences. For purposes of illustration, assume that the average or mean number of errors made by file clerks is two errors per one hundred units of work. While some clerks will make more than two errors in filing one hundred units, and others will make fewer than two errors per one hundred units, the majority will make two errors in filing one hundred units.
The number of errors per total output is important because this factor has a bearing on the number of errors or deviations that can reasonably be expected to occur within a given number of observations. Deviation is the distance from the mean and is calculated by using the formula for standard deviation. In any normal distribution, 68.3 percent of the total will fall within one standard deviation above and below the mean of the distribution; 95.1 percent will fall within three standard deviations above and below the mean.
In the foregoing example, it was found that the file clerks made an average (mean) of two errors per one hundred units filed. When using the appropriate formula to calculate standard deviation, assume that a standard deviation of 0.53 is found. According to the normal distribution, 68.3 percent of the number of errors made by each file clerk will be within one standard deviation above and below the mean, thus ranging from 1.47 to 2.53 (2.0 – 0.53 and 2.0 + 0.53). two standard deviations above and below the mean (from 0.94 to 3.06 errors) will encompass 95.1 percent of the number of errors made by each file clerk. Three standard deviations will encompass 99.7 percent of the errors. Therefore, all but 0.3 percent of the number of the number errors made by each file clerk will be between 0.41 and 3.59 errors per one hundred units filed. This is illustrated in figure 29-1.
Control limits are the third element of statistical quality control. These limits must be established in order to determine at what point the errors are considered to be due to chance and at what point they are attributable to some identifiable cause. For instance, in the foregoing example. If the control limits are set two standard deviations above and below the mean. Chance is the cause of errors that range from 0.94 to 3.06 errors per one hundred units. The chance nature of these errors makes it impossible to identify the reasons employees make such errors. On the other hand, some identifiable cause is responsible when employees have errors exceeding 3.06 per one hundred units. In this particular example, the errors may be due to poorly trained employees. Figure 29-2 illustrates the control limits for this particular example.
With the exception of sample 4 in figure 29-2 all samples fall within the control limits. Therefore, these errors are due to chance. Since the number of errors in sample 4 exceeds the maximum for the upper control limits, some factor other than chance is responsible. One speculation is that the sample was taken at the end of the work day when employees were fatigued.
Zero Defects. The last of the quality control techniques presented in this chapter is zero Defects. The fundamental characteristic of this technique is the prevention of errors because employees do their work correctly the fist time. Thus, there is more to Zero Defects than identification and correction of errors. Employees must also be motivated not to make the errors in the fist place.
In using the Zero Defect technique, employees pledge to management that they will support the concept and will produce error-free work. They also make suggestions to management about ways to eliminate errors. And they are subsequently rewarded on the basis of the quality of the suggestions they submit, on the basis of the improvement they make, and on the basis of their error-free work. In some instances, errors are attributable to management rather than to the employees themselves. If nothing is done to eliminate those that are attributable to management, the effectiveness of the Zero Defects program cannot be fully realized.