Chi-Square Methods
To view active media links, please use the Training Programs tab.
Topic Description
The Chi-Square Methods topic will provide the participant with the knowledge and skills necessary to employ several key forms of the chi-square statistic. Of interest, the chi-square test is any statistical hypothesis test in which the test statistic has a chi-square distribution, given that the null hypothesis is true. This test statistic is often employed to determine the underlying distribution of a product or service performance variable or estimate the extent of association between one categorical variable and another. It is also a foundational statistic when conducting survey-based investigations and research, such as customer satisfaction analyses.
In this sense, the chi-square statistic is a highly versatile tool that has many applications in the commercial and industrial world of business. As such, it represents a foundational tool for anyone involved with quality improvement activities that are dependent upon the analysis of frequencies and proportions, such as comparing the before-and-after defect rate of a product, service or transaction.
Reinforcement of major concepts, techniques, and application is realized through exercises, scenarios, and case studies. The following prerequisite topics are listed in sequential learning order: Basic Statistics and Hypothesis Testing. Total instructional time for this topic is 3 hours and 19 minutes.
- Statistical Definition - Describe how to translate a practical problem into a statistical problem
- Model Fitting - Explain what is meant by the term "Model Fitting" and discuss its practical role in Six Sigma work
- Testing Independence - Explain how a test of independence can be related to the idea of correlation
- Contingency Coefficients - Understand how a contingency coefficient relates to a cross-tabulation table
- Yates Correction - Describe the role of Yates correction in terms of the chi-square statistic
- Testing Proportions - Test the significance of two proportions using the Chi-square statistic