Collecting data for analysis
A data checksheet is any form designed to help collect data quickly and easily. It is also sometimes referred to as a tally sheet or data collection sheet. Data collection is important because it is the starting point for statistical analysis. The more pertinent, accurate and comprehensive the data, the more reliable will be the conclusions drawn from its later analysis. If data are collected using a purpose-designed form which is easy to use, then there is a greater likelihood that accurate information will be gathered.
There are no strict rules for how a particular data checksheet should look. Use the guidelines and examples in this document to design your own.
Use it when you need to get to the facts
Whenever you are engaged in process analysis, you will come across aspects of activity which you are unclear about. For example, when reviewing a process such as "handling returned goods", you may know that the reasons for goods being returned include incorrect deliveries, warranty repairs and goods not to specification, but you may be unclear as to how many goods are returned over time and the relative importance of each type of return. The only way to get a clearer picture of the true situation is to collect information, and this is done by designing a data checksheet for the handlers of returned goods to complete.
How to design a checksheet
For the "handling returned goods" example given above, you may design a checksheet in the format shown below.
Always identify clearly the process being examined, the area or location from which the information is being collected, the way in which the information is being collected, and the name of the person completing the sheet. Often the easiest way to record each instance is by making a `tally mark' or stroke as shown above. Any totalling required can be done at the end of each recording period (in the above example this is at the end of each day).
Different ways to collect data
A thoughtfully designed data checksheet will not only make data collection easier but can often provide a shortcut to analysis. For example, if you are involved in manufacturing and you need to know what the degree of variation is in a particular measurement for a manufactured part, you can use a data checksheet such as the one shown below.
This type of checksheet automatically shows a frequency distribution of the data. This is easier than recording all the measurements first and then doing the analysis. Information can also be collected in pictorial form. The example below is a checksheet for recording defects in the finish on car bumper bars.
You can even design checksheets which allow you to begin tracing the causes of variations in the data you wish to analyse. The checksheet below is in the form of a grid with boxes for different machines and operators. Numbers for different types of problems are marked in boxes for each day of the week.
Other references
Guide to Quality Control, Kaoru Ishikawa, Asian Productivity Organization, 1991
to HCi Services
