What are the earliest accounts of publication bias?

Question

Publication bias is a huge issue is some academic fields. What are the earliest accounts of publication bias?

---

As requested in the comments, publication bias intuitively means that if you read one article showing that drug X works to treat disease Y, you should wonder how many articles showing the same drug has no affect are unpublished due to the habits of publishing only positive results.

Here is a video explaining what publication bias is (15 minutes, worth watching if not familiar with publication bias):

Half of all clinical trials ever completed on the medical treatments currently in use have never been published in the medical literature. Trials with positive results for the test treatment are about twice as likely to be published, and this applies to both academic research and industry studies.

Definition from the Wikipedia page on publication bias:

Publication bias is a bias with regard to what is likely to be published, among what is available to be published. Not all bias is inherently problematic – for instance, a bias against publishing lies is often a desirable bias – but one problematic and much-discussed bias is the tendency of researchers, editors, and pharmaceutical companies to handle the reporting of experimental results that are positive (i.e. showing a significant finding) differently from results that are negative (i.e. supporting the null hypothesis) or inconclusive, leading to a misleading bias in the overall published literature.

Answer 1

Here some early papers, listed in chronological order:

1. Sterling TD (1959) Publication Decisions and Their Possible Effects on Inferences Drawn from Tests of Significance--Or Vice Versa. Journal of the American Statistical Association 54: 30–34.

2. Rosenthal R (1979) The file drawer problem and tolerance for null results. Psychological Bulletin 86: 638–641.

3. Simes RJ (1986) Publication bias: the case for an international registry of clinical trials. J Clin Oncol 4: 1529–1541.

4. Dickersin K (1990) The existence of publication bias and risk factors for its occurrence. JAMA 263: 1385–1389.

5. Chalmers I (1990) Underreporting research is scientific misconduct. JAMA 263: 1405–1408.1. Sterling TD (1959) Publication Decisions and Their Possible Effects on Inferences Drawn from Tests of Significance--Or Vice Versa. Journal of the American Statistical Association 54: 30–34.

Answer 2

Clinical trials is an excellent metaphor explaining this bias issue. The results of any clinical trial are probability curves with generous standard deviations. All clinical trials both fail and succeed in curing disease. Thus bias depends on when the lie is born, which depends on the data analysis performed by an observer. A bias against publishing lies is oxymoronic, and can't be quantified, as all bias is based on lies. Similarly, the first casualty in war is Truth. But when have we never not been at war?