Statistical Inference: The Big Picture

Robert E. Kass

DOI: https://doi.org/10.1214/10-STS337

2011-06-22

Abstract:Statistics has moved beyond the frequentist-Bayesian controversies of the past. Where does this leave our ability to interpret results? I suggest that a philosophy compatible with statistical practice, labeled here statistical pragmatism, serves as a foundation for inference. Statistical pragmatism is inclusive and emphasizes the assumptions that connect statistical models with observed data. I argue that introductory courses often mischaracterize the process of statistical inference and I propose an alternative "big picture" depiction.

Other Statistics,Methodology

What problem does this paper attempt to address?

The main problem this paper attempts to address is the foundational philosophical issue of statistical inference, specifically how to find a method in modern statistical practice that can reconcile the views of both the Frequentist and Bayesian schools. The author, Robert E. Kass, proposes a new philosophical perspective—Statistical Pragmatism—which emphasizes the connection between theoretical assumptions and actual data, rather than focusing excessively on the traditional disputes between the Frequentist and Bayesian schools.

### Core Issues of the Paper:
1. **Philosophical Foundations of Statistical Inference**: The paper explores how statistics can move beyond the past debates between the Frequentist and Bayesian schools and proposes a more practical and inclusive philosophical framework.
2. **Problems in Statistical Education**: The paper points out that current statistical education often fails to accurately reflect the realities of modern statistical practice, particularly in the misleading explanations of the basic concepts of statistical inference.
3. **Connection Between Theoretical Assumptions and Actual Data**: The author emphasizes that, whether from the Frequentist or Bayesian perspective, the key to statistical inference lies in the degree of match between theoretical models and actual data, which is the core viewpoint of Statistical Pragmatism.

### Solutions:
- **Statistical Pragmatism**: This is an open and practical philosophical viewpoint that acknowledges the value of both Frequentist and Bayesian methods but stresses the importance of whether the assumptions behind these methods reasonably describe the variability in actual data.
- **Teaching Recommendations**: The author suggests adopting a more comprehensive perspective in statistical education, emphasizing the connection between theoretical assumptions and actual data, rather than focusing solely on the specific technical details of the Frequentist or Bayesian schools.

### Conclusion:
By proposing Statistical Pragmatism, the paper aims to provide a philosophical foundation for statistics that better meets the needs of modern practice and offers a new framework for statistical education, enabling students to better understand and apply the basic principles of statistical inference.