Using statistics in management
All processes in all industries show variation.
Whether you manufacture left-handed sprockets for widgets, or you work in a service industry, daily life always throws up variations in performance. The skill of management is to manage those variations and keep them within acceptable limits.
To manage properly, therefore, calls for an understanding of variation and, in turn, of statistics. Modern managers are developing a better understanding of ‘statistical process control’, to control and reduce the variability in their processes.
Numbers and data
Numbers and data are not the same thing. Data are almost any kind of information useful in solving problems — and of course it often includes numbers. Numbers alone are often meaningless measures, or counts, which can confuse. Data arising from counts are called discrete data and can only occur at definite points.
For example, there can be 0, 1, 2, 3, 4, etc. mistakes on a page. There cannot be 3.768 mistakes on a page.
Data that result from measurements on a continuous scale are called continuous data or variable data. Some scales are measured in discrete quantities so it is inappropriate to define all continuous data as being measured on a scale; the scale must be continuous. For example, British shoe sizes are measured on a scale, but the scale is discrete. For instance, you may take a size 6, or a 61/2 but you cannot take a 63/4 since such a size does not exist.
There are some basic data-handling techniques that many people at work need to understand and practise.
They include: process flow charting, tally chart monitoring, histograms (bar charts).Pareto analysis.
cause-and-effect analysis, using Ishikawa diagrams, scatter diagrams.control charts.
Flow-charting is a technique for depicting the steps or elements involved in a process of some kind. For example, at school you may have drawn a simple flow chart to show the steps involved in making a cup of tea. In industry, you will need to draw flow charts to describe processes in the workplace.