You got a null result. Will anyone publish it?
Selective publication of only “positive” scientific findings is blamed for wasted effort and the replication crisis, as null or negative results are rarely accepted by high‑impact journals or rewarded in academic careers. Commenters describe how incentives around novelty, citations, and funding push researchers to abandon or hide null results, even though these are crucial for bounding effect sizes, avoiding repeated dead ends, and giving a truer picture of reality. Proposed fixes include dedicated venues for negative findings, preregistered study designs, better use of preprints and data repositories, and rethinking evaluation metrics so that careful null results and replications count as valuable scientific contributions.
Publication bias & incentives
- Many see strong bias toward novel, positive findings, especially at elite journals; null results are rarely accepted, even when well done.
- Career incentives (tenure, funding, h-index) push researchers to prioritize “sexy” results over careful nulls or replications.
- Some argue this systemic behavior is now close to scientific misconduct; others frame it as predictable outcome of misaligned incentives rather than fraud.
Replication crisis & statistics
- With only positive outcomes published, false positives are inevitable and often unrecognized; regression to the mean then makes replications “fail.”
- Multiple commenters stress that one study is just one sample; robust knowledge requires different samples and replication.
- Misunderstandings of p-values and “statistical significance” recur; some note that insignificant results don’t prove the null, and huge samples can make trivial effects “significant.”
- Alternatives like confidence intervals, Bayesian methods, and higher significance thresholds in some fields are mentioned.
Value and risks of null results
- Nulls can bound effect sizes, correct false beliefs, and prevent wasted effort, but not all nulls are equally interesting.
- Concern: trivial or sloppy nulls could flood literature or be used strategically against rivals. Others respond that peer review and low citation impact limit this.
Who should do replications?
- Suggestions: make replication and logging of nulls a standard part of PhD training, or pre-register methods before data collection.
- Pushback: this is seen by some students as drudgery and bad for careers; others say it’s essential training but must be properly funded and recognized.
Alternative venues & formats
- Mention of journals and workshops dedicated to negative or “unsurprising” results, plus arXiv, preprints, blogs, and “living papers.”
- Objection: these outlets often carry less professional weight, so researchers deprioritize them despite personal willingness to share nulls.
Structural and process issues
- Peer review often evaluates results, not just methods; proposals for “results-blind” review and conditional acceptance based on pre-registered protocols.
- Cost and bureaucratic load of publishing, limited replication resources, and opaque reviewing further discourage null-result publication.