A Two-Sided Discussion of Preregistration of NLP Research

TL;DR

This paper critically examines the potential benefits and drawbacks of preregistration in NLP research, highlighting challenges like bias, p-hacking, and reduced risk tolerance, through a balanced dialogue.

Contribution

It provides a nuanced, balanced discussion of the advantages and disadvantages of adopting preregistration in NLP, emphasizing the complexity of implementing it effectively.

Findings

Preregistration can prevent fishing expeditions and promote negative results.

It may introduce biases and increase publication bias.

It could lead to more p-hacking and reduced risk-taking.

Abstract

Van Miltenburg et al. (2021) suggest NLP research should adopt preregistration to prevent fishing expeditions and to promote publication of negative results. At face value, this is a very reasonable suggestion, seemingly solving many methodological problems with NLP research. We discuss pros and cons -- some old, some new: a) Preregistration is challenged by the practice of retrieving hypotheses after the results are known; b) preregistration may bias NLP toward confirmatory research; c) preregistration must allow for reclassification of research as exploratory; d) preregistration may increase publication bias; e) preregistration may increase flag-planting; f) preregistration may increase p-hacking; and finally, g) preregistration may make us less risk tolerant. We cast our discussion as a dialogue, presenting both sides of the debate.