Check Sheet: How to understand it

The Quality Toolbook > Check Sheet > How to understand it

When to use it | How to understand it | Example | How to do it | Practical variations

<-- Previous | Next -->

How to understand it

If data is collected in a disorganized way, it is likely to end up as a jumble of numbers on a convenient scrap of paper, where the numbers are easily misunderstood and the paper may be lost. By collecting it in an organized way, fewer mistakes are likely in the collection, transcription, understanding and storage of the data.

A Check Sheet is simply a sheet of paper organized to simplify and standardize manual data collection and to ease interpretation of results. The simplification is characterized by the use of checks or marks to record events, rather than recording these as numbers or text, as in Fig. 1. This enables one Check Sheet to contain a large number of recorded data points.

Fig. 1. Simplifying and standardizing manual data recording

There are three main uses of Check Sheets. Firstly, they can be used to count items, either as a simple count of different items, such as defects, or to show the distribution of a set of measurements. Simple counts may be displayed as a Pareto Chart, whilst distribution may be displayed as a Histogram. When counting items, the Check Sheet also is useful as the overall picture is built up in front of your eyes as you add individual items.Secondly, they may be used to show the physical location of something, such as defects on a manufactured item. This is useful for finding significant bunching of measurements which may then help to find problems. Lastly, they may be use to help prompt for an action and consequently be ticked to certify that a particular action has been carried out.

In interpreting and analyzing Check Sheets, identified problems may be broken down further by using the other information on the sheets about the circumstances where the measurements were made. A danger in interpretation is in not having or using this information and making assumptions that localized problems are more widespread than they are. For example, a defect log from a single product line may be wrongly generalized to cover all product lines.

<-- Previous | Next -->