Keep your enemies close: Adversarial collaborations will improve behavioral science.

Abstract

Behavioral scientists enjoy vast methodological freedom in how they operationalize theoretical constructs. This freedom may promote creativity in designing laboratory paradigms that shed light on real-world phenomena, but it also enables questionable research practices that undercut our collective credibility. Open Science norms impose some discipline but cannot constrain cherry-picking operational definitions that insulate preferred theories from rejection. All too often scholars conduct performative research to score points instead of engaging each other’s strongest arguments—a pattern that allows contradictory claims to fester unresolved for decades. Adversarial collaborations, which call on disputants to codevelop tests of competing hypotheses, are an efficient method of improving our science’s capacity for self-correction and of promoting intellectual competition that exposes false claims. Although individual researchers are often initially reluctant to participate, the research community would be better served by institutionalizing adversarial collaboration into its peer-review process.

The adversarial collaboration within each of us. Comment on Clark et al.

“…Clark et al (2022) presented a Table 2 in which very broad issues are to be investigated using adversarial collaboration. It is not clear to me that this is the best grounds for such a collaboration.…”

“…Perhaps. My view is that Clark et al (2022) have done a great job of summarizing strengths of adversarial collaborations, and my goal here is complementary. I want to describe an overall approach that puts adversarial collaborations in perspective, and I want to highlight reasons why we need individual investigators' work along with them, enhanced by the collaborations but not constrained by them.…”

“…Given the aforementioned state of affairs, how should editors and reviewers respond toward adversarial collaborations? Clark et al (2022) suggested that "When an article claims to oppose or challenge another hypothesis or theory, editors could insist on adversarial collaborations" (p. 11). I find this to be an overzealous way to go.…”

“…A case has been made (Clark et al, 2022; Cowan et al, 2020) that adversarial collaborations contribute to scientific advancement. If you have already read Cowan et al (2020) which was a description of an extended adversarial collaboration on working memory, you may wonder why I have written this essay without including any of my former co-authors.…”

Adversarial collaborations in behavioral science: Benefits and boundary conditions. Comment on Clark et al.

“…A core motivation for adopting adversarial collaborations is the avoidance of scientific misconduct or questionable practices researchers might engage in when pressured by social motives such as career aspirations (Clark et al, 2022). Although such instances do exist, their prevalence is often overestimated.…”