Data considerations for Cause-and-Effect Diagram

To ensure that your results are valid, consider the following guidelines when you collect data, perform the analysis, and interpret your results.

The effect must be a dependent variable

The response or the output of the process is a dependent variable. For example, the response variable might be the wait time for service. In a cause-and-effect diagram, the dependent variable is the variable (or effect) that is affected or measured.

A separate cause-and-effect diagram is required for each response.

Causes can be independent variables, intermediate variables, or nuisance variables

Independent variables: Independent variables are factors that can be directly controlled or adjusted. For example, Training is an independent variable because managers can directly change the methods used to train employees.
Intermediate variables: Intermediate variables are similar to independent variables. They can be adjusted, but unlike independent variables, they cannot be directly controlled. For example, metal strength may depend on the size of gas bubbles in the material. The manufacturer cannot directly change gas bubble size, but can adjust other factors, such as mixing rates, chemical composition, and curing temperature which affect gas bubble size. Thus, gas bubble size is an intermediate variable.
Nuisance variables: Nuisance variables are factors that cannot be controlled or adjusted, but that affect responses. For example, rain and temperature are nuisance variables because the managers cannot control them.

Brainstorm all possible causes

Brainstorming is a team-based technique for listing potential causes for a problem based on open discussion and exchange of ideas. The more ideas that you generate in a brainstorming session, the more likely you are to identify a solution. It is important not to criticize ideas during the process. No idea is too unrealistic to be listed.

After deciding on the critical categories, the team then uses the cause-and-effect diagram to identify causes within those categories that influence the effect under study.