Contingency Tables

Jay Kay

Hi, Quick question to the experts out there, so in Quality engineering, which I figure is part of the field is "Industrial Statistics" so to say.
I am wondering are contingency tables a tool that is actually used to determine relationships between the row's and columns like for example Machine station and total production output.

I believe the concept is calculating the test statistics using Chi Squared Σ [(observed - expected)^2 / expected] where null is no relationship , alternate is yes relationship. Is this tool actually used in the industry or is this too academic / theoretical? Any feedback appreciated

Bev D

Heretical Statistician
Super Moderator
It is used typically when looking at measurement effectiveness of attributes data. It can be used for your example but there are other ways that are more common for this type of analysis such as Control charts, ANOM or ANOVA or the use of confidence intervals in a GRAPH.

Chi square is not the only statistical test for such tables...

In general, unless you are submitting something to a statistical review body (FDA, USDA...) you are better off understanding how to seelct proper sample sizes and structuring a robust analytical study.

I always suggest a start with Deming's seminal paper: “On Probability as a Basis for Action”, American Statistician, November 1975, Vol. 29, No. 4, pp. 146-152

An easier read is Donald Wheeler's Analyzing Data – The Effect of Variation, Quality Digest March 1996
Top Bottom