Implementation of Correlation Analysis to Work Sampling Results

Work sampling is the process that estimates the proportion of time of the workday spent by workers (or machines) on different activities (work or non-work), based on sufficient random observations of activities over a period of time. When jobs do not have short cycle times or high repetition rates the use of such statistical technique is required, because work sampling data can provide information that can be used to establish standards. The work sampling procedure is well standardized by the recommended steps, but some contributions are possible in the evaluation of the observations. In the paper we present our contribution in the improvement of the decision process based on the work sampling data. We introduce the correlation comparison of the measured hourly proportions of all pairs of activities to check, if there are some interrelationships. The results enable easier decision making (conclusions) about the influence of the selected activities to trigger the others. This is bringing us to better production system's efficiency and productivity.