FreeQuality.org - Chart Selector (Beginner)

- Monitor and Detect Process Variations over time?
         If so, what type of data are you dealing with? Variable or Attribute

 (Please note that variable data is continuously measured for example temperature, time, viscosity,length,  cost and figures. Attribute data often takes the form of counts of discrete events, for example shipping  errors, % waste, absenteeism, yes or no data.)

- Are you interested in detecting small shifts with your control charts? Are your process operations short-  runs and small batches? If so, then we recommend using either a EWMA-chart or a Cusum-chart.

- Trying to Determine appropriate Sample Size?

- Trying to figure out what problems are most critical?

- Identify Cause and Effect of Problems?

- Do you have Multi-variable Data? Need to Decide how many design of Experiments to use?

Variable continued.. 

Are your subgroup averages easily calculated? If so proceed to the following question.
 
If not is your answer then we currently do not have a tool that can assist you. We recommend using a median control chart.

Do you have a small sample size (less than 9)? If yes then, we recommend you use the X-bar and R-bar charts.

If no, then we currently do not have a tool that can assist you. We recommend using a median control chart.

Attribute continued..

Are you measuring the number of actual defective parts? For example, number of failed units regardless of how many actual defects they contain.

If your answer is yes, then you are measuring defective data and must next establish if the sample size is constant. If you sample size is constant and is greater than or equal to 50, we recommend you us the NP-chart.

If you sample size is not constant and greater than or equal to 50, then we recommend using a P-chart. 

If you answer no, then again you are measuring defective data. A defect is a unit that fails to meet one of the  acceptance criteria (i.e. measuring only one defect per unit) and must determine if you sample size is constant.

If you sample size is constant and greater than 5, we recommend using a C-chart. If your sample size is not constant, we recommend using a
U-chart.