The Same, Similar, or Different? The Probability of Agreement Methodology for Evaluating Practical Equivalence
Applying agreement methodology to your data exploration
Two-group comparisons exist in any field for which data is used to aid in decision-making. Traditionally, such comparisons are carried out using hypothesis tests with the null assumption that the characteristics under comparison are identical. With sufficient evidence, one rejects this null assumption, and with insufficient evidence one does not. However, simply rejecting or failing to reject the equivalence of these characteristics may not adequately address the question investigators hope to answer.
In this Statistically Speaking, Nathaniel Stevens emphasizes the utility of collecting evidence in favor of equality rather than against it, providing an overview of the probability of agreement framework for evaluating data in this way. By illustrating this concept with applications such as measurement system comparison, reliability analysis and response surface optimization – among others – Stevens makes the case for adopting agreement methodology.
Q&A with the speaker is available at the conclusion of the keynote address.
What you’ll learn:
- Why practical equivalence among characteristics in two-group comparisons is relevant.
- How applications of agreement methodology can enhance data analysis.
- An overview of the probability of agreement framework using various applications.