Check Sheets

A Check Sheet is one of the Seven (7) Basic Quality Tools. A Check Sheet is a simple tool for data collection. In this post, I’ll show how to create and use a Check Sheet and provide some basic heuristics and common pitfalls of Check Sheet and data collection overall. In this article and 3:53 minute HD video, you will learn what a Check Sheet is, what it used for, how to create one, see various examples of check sheets. Download a Check Sheet template.

A Check Sheet allows us to use a structured & standardized format for collecting defect information:

• Minimizes the chance that different people will collect the data in the different ways
• Forces the team to consciously think about what data they want to collect, why they want to collect it, and what they plan to do with the data once collected

What is a Check Sheet

• Check sheets are data collection forms constructed to record specific information about a process in a organized manner
• Check sheets are designed to answer the question How often does this event occur and under what conditions?
• Check sheets can be used either to support conclusions concerning a process on its own or may be a step in the process of assembling the data to support one of the other improvement tools

Elements of a Check Sheet

• A description of what data is being collected need to make sure that clear Operational Definitions are established for each variable being collected
• Places to put the data (actual ease of data collection / recording needs to be as simple as possible tally sheets work well)
• Room for comments both for the defect types and general (in case the data either does not fall into one of the pre-set categories or in case something special happens e.g. a power failure – that needs to be documented)
• Room for potential statistical data

Check Sheets – Key Points

• Keep the form simple to use and understand; (the easier you make the sheet to use, the more likely you will get the data you need)
• Include only information that you intend to use
• Pilot or test the Check Sheet form before you use it on a full-scale basis – Make changes to it if necessary

How To Create a Check Sheet

• As a team, brainstorm to determine the exact data to be collected and how this data will be organized into categories and then subcategories (taking the time to really think this through will greatly aid later on in the project)
• Always stay focused on the improvement opportunity (what are you trying to learn about) and keep asking yourself what you will use the data for
• Develop exact Operational Definitions of each variable
• Construct the data collection form or forms that will allow recording of the observations or events in their proper classifications as determined earlier
• Test a draft copy of the check sheet by confirming that all possible observations or events may be recorded in one and only one place on the check sheet
• Make necessary corrections to the check sheet
• Collect the desired process data using the check sheet

Check Sheets – Points to Remember

• Allow extra space on the check sheet to record important observations not anticipated during the design of the check sheet
• Determine and record the time period and other process conditions that will be fixed or constant during the observation
• Make sure that the data form and the meaning of all the labels are understood equally and consistently by all team members
• Allow enough time for all observations to be correctly made and recorded during the collection process
• Remember, a check sheet provides only organized raw data. It may be necessary to use another improvement tool to better interpret the results.

Check Sheets – Common Pitfalls

• Measuring things that do not answer the question being asked
• Inability to obtain sufficient sample size
• Associates simply not taking measurements or collecting data (sometimes as a result of strong focus on making their numbers)
• Lost check sheets should keep track of how many are handed out and how many come back
• We can get that from the system attitude resulting in spending more time trying to query log files than it would take to manually collect the data
• Poor Operational Definitions resulting in large amounts of noise in the data