Trustworthiness of statistical inference

TL;DR

This paper discusses the importance of trustworthiness in statistical inference, defending traditional tools like p-values against criticism, and emphasizes proper use and understanding to maintain trust in science.

Contribution

It argues that trust in statistical inference relies on the trustworthy use of tools like p-values, and criticizes calls to ban them, proposing alternative solutions to improve understanding and trust.

Findings

Misuse of statistical tools damages trust in science.

Banning certain statistical methods may harm scientific progress.

Proper education and interpretation can enhance trust in statistical inference.

Abstract

We examine the role of trustworthiness and trust in statistical inference, arguing that it is the extent of trustworthiness in inferential statistical tools which enables trust in the conclusions. Certain tools, such as the p-value and significance test, have recently come under renewed criticism, with some arguing that they damage trust in statistics. We argue the contrary, beginning from the position that the central role of these methods is to form the basis for trusted conclusions in the face of uncertainty in the data, and noting that it is the misuse and misunderstanding of these tools which damages trustworthiness and hence trust. We go on to argue that recent calls to ban these tools would tackle the symptom, not the cause, and themselves risk damaging the capability of science to advance, and feeding into public suspicion of the discipline of statistics. The consequence could be aggravated mistrust of our discipline and of science more generally. In short, the very proposals could work in quite the contrary direction from that intended. We make some alternative proposals for tackling the misuse and misunderstanding of these methods, and for how trust in our discipline might be promoted.