Abigayle's Writing Portfolio

Click here to view or download this document using Word.

Abstract

The purpose of this investigation was to practice work sampling using statistics and occurrence sampling. This was investigated by making observations to determine the probability of occurrences of specific events. After reviewing these observations we found the median and most frequent occurrences, calculated a sample size and used a computer to create randomized times to take the observations. The results showed that choosing an appropriate sample size and being able to generate random times to take the observations is a complex process.

Introduction

It has been suggested that the sample size and the times the samples are taken are important parts of gathering accurate data. This information is important because without good data and knowledge of problem occurrences, a designer can neither fix or improve the process. In order to assure that the information is acceptable a good plan is necessary as well as the ability to use the information gathered wisely. In this experiment, an administrative assistant was observed doing 32 activities and the frequency of these activities were analyzed.

Materials and Methods

The materials necessary to complete this experiment include: a Occurrence Sampling Sheet and the computer program ErgoMaster.

This experiment entails the following procedure:

1. Observe the worker doing 32 activities and record the activities according to the category the activity falls under.
2. Calculate the probability for each category.
3. Calculate the sample size for the categories with the most frequency of occurrence and the median frequency of occurrence using 95% confidence and 5% error.
4. Use ErgoMaster to generate sampling times assuming 100 samples will be taken, there are 9 categories, samples will be taken during the first shift on Monday which lasts from 7am to 5pm with one 1-hour break lasting from 12pm to 1pm.

Results

After 32 samples were taken, “PC Use” was the category that saw the greatest frequency of occurrence (9 out of 32) and the median occurrences were “Fax/Copy in Office” and “Copy Room” with 3 occurrences each. The desired sample size for PC Use was 3914 and for “Fax/Copy in Office” and “Copy Room” were 14700 each. For all results, see Appendix A. The random sample times found are listed in Appendix B.

Discussion

The sample size of an observation is an important part of any data collection process. There are several things to consider when calculating the sample size. First, the frequency of the activity being observed should be considered. The observer must decide whether to use the most frequent or median frequency occurrence to calculate sample size. Using the median frequency occurrence can often be best. After comparing the most and median, it is obvious that the amount of information gathered will not vary excessively. If the problem occurs frequently enough, the results for the median frequency will have very similar information so collecting more on the same day is unnecessary. Although every problem may not be seen by using a sample size based on the most frequent, most of these problems are not severe enough to warrant an increase in sample size. So, by using the median frequency occurrence sufficient information is gathered about problems that occur often enough to require an adjustment, while at the same time not gathering so much information that data is repetitive. Finally it is important to have computer generated random observation times to avoid the pitfalls of periodicity.

Click here to view or download this document using Word.