Teaching: confidence, prediction and tolerance intervals in scientific practice: a tutorial on binary variables.

Area of Science:

• Epidemiology
• Biostatistics
• Statistical Modeling

Background:

• Interval estimates, including confidence (CI), prediction (PI), and tolerance (TI) intervals, are increasingly important in epidemiology.
• While CIs are common, PIs and TIs are less familiar to researchers.
• Understanding these interval types is crucial for robust statistical practice and greater knowledge gain over point estimates.

Purpose of the Study:

• To provide a concise, hands-on tutorial on two-sided CI, PI, and TI for binary variables.
• To explain the meaning and applicability of these intervals from both classical and Bayesian viewpoints.
• To offer practical guidance and R code for direct application in research.

Main Methods:

• Tutorial based on existing teaching materials.
• Explanation of CI, PI, and TI concepts and their uses.
• Demonstration using a worked-out example from veterinary medicine.

Main Results:

• Provides clear explanations of CI, PI, and TI for binary data.
• Offers practical R code for implementing these interval estimates.
• Illustrates application with a real-world veterinary medicine example.

Conclusions:

• The tutorial serves as a valuable resource for teaching statistical intervals.
• Suitable for both classroom instruction and self-study by students and researchers.
• Aims to improve the understanding and application of CI, PI, and TI in practice.