1
SciPoll 640: Impact of publication bias
In your field, how prevalent do you believe publication bias is?
Results
(203 Answers)
Answer Explanations
- Somewhat prevalentuser-725842I believe publication bias is somewhat prevalent in the biotech and pharmaceutical industry. The pressure to produce positive results can lead to selective reporting, where studies with favourable outcomes are more likely to be published. This skews the scientific literature, overrepresenting successful studies and underreporting null or negative findings. As a result, systematic reviews and meta-analyses may be compromised, potentially leading to inflated efficacy estimates and misleading conclusions. Addressing this bias is crucial for maintaining the integrity and validity of scientific research, ensuring that the full spectrum of findings is represented and considered.
- Very prevalentuser-575039Publication bias is very prevalent in my work setup. In my professional experience most of us did not give value for research findings with no positive result and with specific association.
- Somewhat prevalentuser-525512publication bias is quite a common problem in the aspects of computer science and engineering. It means statistically significant or results matching preliminary expectations are more probably published while studies returning null or results that are non-significant often go unnoticed or unpublished. This can result in literature that is overrepresented with positive or significant findings, thus giving a misrepresented view of the real body of knowledge. Publication bias results from a mixture of factors, mainly the preference for novel and exciting results, pressure to publish positive findings, and limited space in journals. Publication bias distorts the complete view that research is supposed to provide with unbiased outcomes and may help in the avoidance of possible research waste.
- Very prevalentuser-111275The majority of publications is not blinded and small number of cases
- Very prevalentuser-350867really problematic in tox. only studies where a chemical is toxic get published. So how do we choose safer chemicals if we do not see the data?
- Very prevalentuser-887652So prevalent that research questions are routinely structured in order to generate positive results.
- Not very prevalentuser-971114I do not believe publication is bias, since the decision is ususally taking by the journal editorial board after thorough evaluation
- Somewhat prevalentuser-610359Lack of awareness and attitude to handle negative results.
- Very prevalentuser-787588In my field, cancer research, reports most frequently show positive results and only minor negative ones when the latter support the positive, more important results.
- Very prevalentuser-887682Journals are always looking some improvement in methodology or for enhanced properties. All journals in our field reject the manuscript if the properties enhancement will be negligible!
- Very prevalentuser-208008Negative studies are rarely presented and often publications just need some data to share, even if it isn't practical or relevant.
- Somewhat prevalentuser-280167.
- Somewhat prevalentuser-110809There's a general view that only positive results are publishable - unless the topic is controversial.
- Very prevalentuser-232888In studies on the efficacy of new drugs for a particular condition, enrolling patients of a specific age group or with very mild symptoms will often lead to a non-generalized result for broader populations, therefore, the real effectiveness of the drug might be overestimated.
- Very prevalentuser-731929It is very rare to find publications with negative results
- Somewhat prevalentuser-902950Although there are quite a few studies on the lack of effect X has on molecule/system/process Y, the majority still centers on positive effects (i.e. some effect was observed in comparison to...)
- Somewhat prevalentuser-765807To mitigate publication bias, efforts like preregistration, open science practices, and transparent reporting are crucial. Researchers and journals play a vital role in addressing this issue.
- Very prevalentuser-640071In the field of biological sciences biases in the publication are prevalent at various levels starting from the submission of an article (so many paid journals claim that payment doesn't impact acceptance of the article but it does, and many journals will not allow you to submit without assurance of payments), reviewing (it is mostly biased towards authors from different countries and regions), acceptance (big names in the area of the research are often given priority over new authors irrespective of the quality of publication), and there is a preferential tendency to accept those publications which are proving of some concept given by peers in the area of research.
- Somewhat prevalentuser-559006Publication of negative results not valued.
- Very prevalentuser-293647I just had a paper rejected as one reviewer said the data were not showing anything so why is this interesting or worthy of publication. From a product stewardship position, having negative findings is very important in the use of a product when determining risk. I feel it is alarming that negative studies are not considered appropriate for publication.
- Very prevalentuser-231488Publication bias occurred not because of publication itself but also the study itself.
- Very prevalentuser-633316Especially as Turkish scientist, I met with this situation in many times. In most cases, it has been raised to almost a racism and my manuscripts receted directly by editors. Mostly reason was out of interest. But before and after my submissiom, many publications published in similar areas by same country scientists of editors.
- Somewhat prevalentuser-149708General lack of interest in null or negative results. By training and cultural bias, we're leaning toward the valorization of the positive results.
- Very prevalentuser-456425IT is based on both Editor perception and Author tendency to not focus on research with tentative negative results.
- Very prevalentuser-570468Journals aren't very keen to accept negative results or replication studies in different geographical areas.
- Very prevalentuser-964888Indian authors rejected.
- Very prevalentuser-307024Almost 60-70% trust contents of published studies.
- Very prevalentuser-781389molecular paleontology is biased by the preconception that molecules have a time limit--first 1000 years, then 100,000, then 1million years, now 3 million years. if you propose, based on data, that some informative molecules may persist over geological time, bias favors conventional wisdom over hard data.
- Not very prevalentuser-433580Being done as per guide lines
- Not at all prevalentuser-109201Publication gives you additional weightage in your career advancement.
- Somewhat prevalentuser-673264Reviewers may tend to adjust their acceptance criteria depending on the reputation of the authors, or the country of origin.
- Very prevalentuser-650602the need to publish sometimes overwhelms the actual science.
And not just any publication, high impact or it doesn't matter. - Somewhat prevalentuser-149459Publication bias is a significant concern in many scientific fields, including pediatrics. Researchers might selectively report outcomes that show significant results, leaving out those that are non-significant. This selective reporting can lead to an incomplete and biased understanding of the research topic.
- Very prevalentuser-665841Always thought we needed a journal of things that didn’t work.
- Somewhat prevalentuser-8496An additional concern is reviewer bias of submitted manuscripts. Reviewers with a specific political or regulatory bias have influenced the acceptance or rejection of a manuscript
- Somewhat prevalentuser-583633The essence of good science is a well thought through hypothesis, rigorous experimental design protocol and an ability to interprétâtes data unbiasedly.
This should lead to a possible successful outcome and published work. This is a fundamental premise of modern science, hence the « bias » towards publishing successful experiments and proving a hypothesis. Often times a poorly thought through hypothesis, lack of understanding of the particular of focus and poor experimenta design and execution leads to a failed corroboration of a hypothesis. Such « papers » should not be published. There are more efficient ways to teach young scientists ( or older ones) about science and how to practice it.
There are cases where a good working hypothesis, rigorous experimental design and execution are carried out, but for a variety of reasons does not work. Such approaches should be published but with the caveat that paper provides new insights into why the experiments/hypothesis failed. - Not very prevalentuser-619625The most important thing is seeking the truth and having a better understanding of knowledge, not the quantity of publications or publishing only positive results
- Somewhat prevalentuser-447943I have experienced and witnessed others experience acts of discrimination due to the place where we carry out our work. Moreover, it is becoming increasingly difficult for us to publish due to the high costs that we cannot afford with the research grants we receive.
- Very prevalentuser-65167basically, negative results are not published
- Somewhat prevalentuser-520983My field is toxicology. It is very hard for a scientist to
- Somewhat prevalentuser-686978I consider that in the biological area, negative results are not interesting in being published in journals.
- Very prevalentuser-388091There are major players, I call "ladder climbers" in major zip codes that push their models - this type of self-promotion works, so their models move ahead motivated by moving ahead. Meanwhile pure scientists, I call "artists", push on, and eventually correct the bias when their data becomes overwhelming and the truth pushes through. Then, the ladder climbers attempt to take credit for the new model, and so on. It all works out in the end, because unlike popular opinion, truth is objective.
- Not very prevalentuser-861631Although publication bias is a concern in many research areas, it is not very prevalent in the field of road studies. This is because projects and research related to roads typically involve quantitative and observable data, such as traffic patterns, pavement conditions, and environmental impacts, which are collected and analyzed rigorously. Additionally, much of the research is conducted by government agencies and independent organizations that follow strict guidelines for transparency and disclosure. As a result, studies tend to be published based on their relevance and methodological quality, regardless of the results found.
- Very prevalentuser-414245There is bias based on which university the paper is coming from. Smaller, state universities, are discriminated against, as are authors. Authors with foreign names, particularly from the eastern countries, are discriminated against. Prejudice is prevalent because of the anonymous nature of the peer review system. This bias is, in my opinion, far more important than the possible rejection of null result studies.
- Somewhat prevalentuser-951310Recently, I have witnessed a main author admit to looking for a lower number to update a toxicity factor, so the lower value was achieved by ignoring toxicological principles of dose-response and what constitute critical effects.
- Somewhat prevalentuser-340576There is a strong pressure to support all results with molecular tools, which is somewhat unnecessary and expensive for some countries.
- Very prevalentuser-778496Having submitted (and published) many journal manuscripts, I have witnessed the poor quality of reviews and monumentally poor decisions, suggesting the presence of extreme bias.
- Very prevalentuser-773118I believe publication bias is very prevalent and actually represnted by the mainstream of academia. That is why eventually the majority of published researches are found unreliable although positive results and good statistics are shown.
- Somewhat prevalentuser-683654People struggle to publish just to be promoted.
- Somewhat prevalentuser-606148A great deal of empirical evidence involves assessment of learning gains from learning environments that include information and communication technology (ICT). These studies are almost always quasi-experimental, and thus lack randomization of the treatment and control groups. Many studies use weak designs. This makes it difficult to be sure that a weak learning gain is actually due to ineffective design or use of the technology application. Furthermore, researchers often have a vested interest in the ICT application they are studying, so they tend to rationalize weak result as due to flaws in the study other than the ICT application. Consequently, studies with weak or negative outcomes are rarely submitted or accepted for refereed publication. If they see the light of day at all, it may be only in a conference proceeding, or just a research archive.
In fairness, even large-scale well funded RCT studies in many aspects of education have led to inconclusive results -- to the point where some leading researchers (e.g, Berliner) have argued that classic experimental designs are not up to the task of analyzing what's going on in a learning environment. In education, Berliner argues, the interaction effects tend to be larger than the main effects. Classical experimental designs have a difficult time in this situation. Furthermore, some researchers, as a result, have argued that the research strategy should trade away some internal validity in order to get more external validity, so they advocate for various multivariate analyses, and qualitative study designs. My opinion is that these designs do often yield a better understanding of the learning environment, but often make it difficult to risk a claim of a negative result. - Not very prevalentuser-50238Through publications research community share ideas, innovations and problems faced.
- Very prevalentuser-169864Negative results are almost impossible to be published. In practice, a negative result requires a huge sample size (the entire population or so)
- Not very prevalentuser-165077In general, good work gets published
- Somewhat prevalentuser-649046The publication bias in favour of positive results is more likley and therefore more prevalent due to inherent nature of how scientific journals functions that largely look for and highlight some “find” rather than none or negative results from research study.
- Somewhat prevalentuser-393430People often highlight their own positive findings and tend to omit (or decrease the importance) conflicting data obtained by others with different approaches. Also personal negative data are rarely mentioned
- Somewhat prevalentuser-625143as always cool findings are more attractive than negative findings. In veterinary medicine case numbers are small so that it is easy to not reach significance.
- Somewhat prevalentuser-734516Some studies, mostly including those with no significant results, are not published. Firstly, due to the lack of interest by the authors and, somehow, by the Journals.
- Very prevalentuser-561710it's prevalent. nothing else to add.
- Very prevalentuser-415326Heavy bias towards publishing trials that are positive and in areas of novelty and specific interest. This leads to unpublished null/negative or equivocal outcomes and a lack of external validation.
- Somewhat prevalentuser-523578Editorial boards promote friends as colleagues and ideas that they agree with.
- Somewhat prevalentuser-56501Completing lab interests with alternative finidings..."keep the other guy's opinion out of lime light"
- Not very prevalentuser-82487With biased publications science will never grow. Since scientific research is growing rapidly, these small temporary hurdles must be ignored
- Very prevalentuser-15550Only "positive" results data is published. Because of that, not all aspects of a subject are considered when we publish
- Very prevalentuser-74194I believe that selection of data and statistical methods to reach a pre-determined conclusion is an ongoing issue in scientific publishing