Year: 2013
Type: editorial
Abstract: In experimental science, nature is the ultimate judge, and this judge is independent, unbiased, and with no vested interest in any particular outcome. Hence, one might argue that if a result is honestly reported, it must be true. True it may be, but what if its truth is ephemeral, observed only under stringent conditions, in one laboratory perhaps, and perhaps only for a short period of time? Then, it may still stimulate others to change their research approach and perhaps hit upon a major discovery, but more often, its value is limited and it may be destined to the dustbin of irreproducible results. Scientific journals have a great responsibility to sift through the submissions and weed out those that offer little in value to the community. In this way, they prevent extraordinary expenses in time and money in attempts to replicate the original findings. We at PCMR take this responsibility seriously, knowing full well, however, that perfection is hard to achieve and that on occasion we fail. It is in this context that we intend to make some changes to the instructions to authors over the next 6 months. A goal of this process is to help ensure that what was true at the time of publication remains so for years to come. The goal is not to make the process of preparing manuscripts more difficult, but rather to clarify the expectations for publication in PCMR for authors as well as reviewers. In its most dramatic form, incorrect or irreproducible data can lead to manuscript retraction. The percentage of manuscripts retracted has steadily increased over the last 30 yr, and retraction is most often due to fraud or other forms of scientific misconduct. It is perhaps no surprise that the frequency of retractions is positively correlated with impact factor. With current computer software, plagiarism is easier to detect, and all manuscripts considered for publication in PCMR are evaluated for plagiarism. However, fabrication of data and other forms of fraud can be very difficult to detect on peer review. A milder form of incorrect data or conclusions is based on unintentional errors. These are more likely to be caught before publication, a major benefit of the peer review process. A simple guideline is that all conclusions made in a particular manuscript should be supported by data that are evaluable and acceptable to an expert capable of evaluating it. Occasionally, the range of types of data in a particular manuscript is beyond the expertise of highly qualified reviewers. In these instances, the PCMR editorial staff makes a strong effort to ensure that additional qualified reviewers with the specific expertise are selected. Manuscripts in which histological data are the basis for a critical conclusion are an example of an area that is typically outside of the expertise of many peer reviewers. This is readily apparent when attempting to evaluate data or histological images in the highest impact journals relative to specialty journals that are more familiar with this type of data. One of the most frequent examples of overinterpretation of data observed in manuscript submissions to PCMR is the presentation of a finding in one cell line. While the finding may be revolutionary if present in a significant fraction of lines, it may not be so if only present as an isolated observation. Hence, we will generally stress that some evaluation of the frequency or generality of a particular observation is required for the publication of original manuscripts in PCMR. The appropriate use of statistics in the evaluation of scientific data is nearly universally required for publication and will be formalized to some extent in revised instructions to authors. However, countless examples of statistically significant differences in artifactual results exist. These generally do not qualify as either fraud or error, as the data are reproducible if the experiments are performed in exactly the same manner as published. The conclusions reached by readers of such data frequently serve as the basis for additional or new experiments that are slightly different and produce data that do not support the original conclusion. However, in this way, the field moves on and scientific knowledge is built up. If several independent laboratories cannot reproduce the core data, it is reasonable to question whether it should have been published in the first place, although we accept that in most cases, these data were published in good faith by the original authors. While PCMR will make efforts to ensure that published data are correct so that the field does not take one step back for every two steps forward, we depend heavily on the integrity of authors to ensure that their work is accurate and representative. In addition, we rely on reviewers to ensure that submitted data and its interpretation are as rigorous as possible. In these ways, we will try to ensure that all work published in PCMR is rigorous and withstands the test of time. Ultimately, nature will be the final arbiter of where we have succeeded in this quest.
Source: Pigment Cell & Melanoma Research
DOI: 10.1111/pcmr.12071
Language: English
Subfield: History and Philosophy of Science
Field: Arts and Humanities
Domain: Social Sciences
Open Access status: bronze
OpenAlex ID: W2076984990