How to Choose a Control Chart Using a Decision Tree

Many people struggle with choosing the right control chart.

A control chart decision tree can help you choose the right control chart based on the data (variable or attribute) and sample size.

 

A Control Chart Decision Tree is Coded into QI Macros Control Chart Wizard

Download a pdf version of this Control Chart Decision Tree

To select the right control chart:

1. Download a QI Macros 30 day trial and let the Control Chart Wizard do the work for you
2. Use the decision tree above
3. Use the question and answer format below to select the right chart.

---

Select the Right Control Chart - Attribute vs Variable Data

• Is the data for the control chart variable data?
  (e.g., time, height, length, width, weight, money?) Yes No
• Is the data for the control chart discrete or attribute data?
  (e.g., number of people, complaints, defects, missed commitments, repair reports, etc.)? Yes No

Control Charts for Variable Data

• Do you collect a single sample of the data on a periodic basis?
  (e.g., your weight on a daily basis? Use the XmR Chart
• Do you collect 2-10 samples of the data at the same time?
  (e.g., length of five concurrent customer calls in a 60 person office? Use the XbarR Chart
• Do you collect more than 10 samples of the data at the same time? Use the XbarS Chart

Control Charts for Attribute Data

• Do you count the number of defective products or services?
  (e.g., number of defective Caller ID units or number of missed commitments?) Yes No
• Do you count the number of defects in a given product or service?
  (e.g., defective splices in a cable or the number of mistakes in a typed letter?) Yes No

Number of Defective Units

• Is the sample size (i.e., the number checked or tested) constant?
  (e.g., missed commitments in a sample of 100 installations/day? Use the "np" Chart
• Does the sample size vary from day to day?
  (e.g., number of incorrect payment to vendors per day? Use the "p" Chart

Number of Defects Per Unit

• Is the sample size (i.e., the number checked or tested) constant?
  (e.g., number of billing errors in a sample of 1000 bills/day? Use the "c" Chart
• Does the sample size vary from day to day?
  (e.g., number defects in all service orders? Use the "u" Chart

---

Why Choose QI Macros Over Other Control Chart Software?