Control sheet – how to collect data?

Control sheet – application


In earlier entries, a lot of speech was about methods of solving problems, from the simplest ones – to be applied in a working group, to more advanced ones – such as analysis of control charts.

Do you already know these methods? If not yet, I invite you to read the previous entries:

DMAIC, Pareto analysis, brainstorming, method 6-3-5, control charts, histogram, 5 Why, cause and effect diagram.


Before starting any kind of analysis of problems, we need one – data from the process! Based on them, discussions, prepared charts and control charts will be conducted, and then actions will be taken.


The data should be:

– current (but sometimes historical)

– reliable

– precise

– collected in a strictly defined way

– collected on the basis of the same sample size, if they are to be compared

– from a specific time interval


The control sheet is a tool that is used to collect data. It is therefore most useful at the beginning of the problem solving process, but also when implementing new processes and products.


The control sheet enables systematic monitoring of process performance, accounting and collating current as well as historical data with each other. Thanks to this we have a full picture of the dominant tendencies and trends in the analyzed process.


What gives us the use of control sheets? Top 5 advantages


– Collecting data from the process in an orderly, readable and easy way.

– A simple and universal procedure for using the spreadsheet facilitates application in any area of the organization’s operations (production, administration, finances, personnel department, purchases, sales, etc.).


Building an image of the condition of a given process “in real time”, on an ongoing basis, based on the collected facts, and not the opinions of those involved in the process.

Observation of trend  that are shaped on the basis of current and historical data.


– Comprehensive application as a starting base for each method of solving problems in the area of quality and continuous improvement.


What should we know to develop a control sheet well?


There is no one way to develop a control sheet. Everything depends on what it is to be used for, what area and what data should be collected using it.

However, it is worth following some basic rules not to overlook some important values.


1. Determine what data you will collect using the control sheet.


This is a specific definition of the observed phenomenon, situation, parameter.


If you are observing the intensity of sickness absence in specific areas of the organization’s activity, within a certain time interval, you must define:

– what is meant by “sickness absence”: eg absence from work due to illness, lasting longer than 1 working day and confirmed by sick leave.


– in what time perspective data will be collected: e.g. week to week in the period from September to December

– what areas will be the analysis: for example, administration and sales staff

Tip: Make sure that everyone who prepares the analysis for this problem uses the same control sheet with the same problem definition. Then you will be sure that the collected data can be combined in any way, compare and draw conclusions.


2. Determine who and how it will collect data


Specifically, a certain standard of conduct must be developed for the collection of data and the data sheet itself – that is, the control sheet. This is especially important when a larger group of people is dealing with the topic – everyone in the same way must understand the definition of the problem and when and how to record observations.

Observations can cover the whole area or only a certain, specific sample, e.g. people employed for less than two years.


No less important is determining whether those responsible for collecting data have sufficient knowledge, how to do it and whether they have a planned time. There is nothing worse in the case of solving a problem / problem, such as haste and inaccuracy at the first stage, that is, collecting information about the process.


Before you proceed to the action, try to collect preliminary information about the problem, to assess how big its scale is and when you can take measurements. It’s about checking if the process is available only at a specific time, or you can start collecting data at any time. On this basis, it will be easier to plan for how long to gather information to collect a representative sample, taking into account all typical and unusual phenomena.



3. Develop a control sheet tailored to the needs of the observed process / phenomenon.


What elements should be included in every good control sheet?


Source data:

What are we testing? – information about what is being observed, i.e. what parameter, size, phenomenon

e.g. sick leave


Problem definition: sick absence – absence from work for more than 1 working day due to illness, which is confirmed by a sick leave.

Where? – a place of data collection, e.g. administration department, employees employed under 2 years

Who? – responsible for collecting data, e.g. a human resources specialist from the personnel department

When? – date of the results, e.g. Monday, Tuesday, Wednesday, … CW20



Examples of control sheets




The method of using such a sheet is extremely simple and requires no specialist knowledge. It is enough to indicate the fact of occurrence of a given phenomenon.

In this case, an example of sickness absence analysis is provided. However, it can be equally well used on the production to monitor eg losses.

Then, instead of CW (current week) and day, you can specify eg the most frequent losses, and instead of the number of occurrences – mark the loss of time in minutes, hours, days, …


  • Control sheet for collecting measurement data – usually used in parallel with the histogram and control cards


Control sheet for collecting measurement data


This example in turn refers to production, and the data is collected by one person. You can also set at every change of the employee responsible for measuring and registering results. However, the more people make measurements that are to be used for one analysis, the greater the risk of interference. It may turn out that the dominant volatility demonstrated is represented by a change to a change, and the actual problem lies somewhere else.



If you are interested in this topic and would like to have access to the example control sheets in an editable form – subscribe to the newsletter!



Leave a Reply

Your email address will not be published. Required fields are marked *