I basically agree with Dr. Subhash Chandra, as a mentor I tell my students that experiments never lie. We are just not smart enough at the time to understand. That is the reason all experiments should be designed using simple "Hypothesis" and test your hypothesis to see if it is correct or not. So experimental design becomes critical to test your hypothesis (and not to prove your hypothesis). Once you have thoroughly analyzed the data and in context of all of the published information, you have the option to modify your hypothesis.

One of my mentors Dr. Dieter Burger told me that "Purity" is always n+1. n = number of ways you tested and 1 = a method you did not use, forgot or you did not know. Science is always evolving and trying to be perfect. Therefore, we just have to see that all known science is flawed so we keep testing, retesting till we get to a point we approach "Purity, Theory, Truth" you name it.

I agree there are several journals that publish "not reproducible Results", but that does not make sense. So we should publish in regular journals.

Hi,

Journals that are trying to publish negative results do exist but unfortunately, their success is limited. Here are some examples:

http://www.jnr-eeb.org/index.php/jnr

http://www.jnrbm.com/

http://www.pnrjournal.com/

Very good idea.

That sounds like a good idea! We need to have a clear definition of "Negative Results". In my view, if a scientifically sound research methodology is followed, and from that the results obtain do not agree with what most other studies report, this may have at least two explanations. FIRST may be that the results from other studies are WRONG - perhaps arising from lack of soundness of their methodology. SECOND may be that the methodology is sound, but the results differ perhaps because of difference in "context".WRONG results do not qualify to be called "negative" - they are WRONG because of lack of sound methodology.    

very nice idea.

It should be. I would like to be a part of such journal.

Dear Dr

thank you for your questions.

It is a good idea. But as you mentioned I think it may be some times experimental errors and this will open the place to all researcher that have a negative results to  publish their work but their is an experimental errors may be as you mentioned before and other error not related to experimental one and you could have a look in any books for Experimental design and statistical analysis. Sometimes the researcher used non suitable statistical analysis.

It need suggestions

with best regards

Khaled 

I basically agree with Dr. Subhash Chandra, as a mentor I tell my students that experiments never lie. We are just not smart enough at the time to understand. That is the reason all experiments should be designed using simple "Hypothesis" and test your hypothesis to see if it is correct or not. So experimental design becomes critical to test your hypothesis (and not to prove your hypothesis). Once you have thoroughly analyzed the data and in context of all of the published information, you have the option to modify your hypothesis.

One of my mentors Dr. Dieter Burger told me that "Purity" is always n+1. n = number of ways you tested and 1 = a method you did not use, forgot or you did not know. Science is always evolving and trying to be perfect. Therefore, we just have to see that all known science is flawed so we keep testing, retesting till we get to a point we approach "Purity, Theory, Truth" you name it.

I agree there are several journals that publish "not reproducible Results", but that does not make sense. So we should publish in regular journals.

Making negative results public is an important part of addressing issues of reproducibility and efficient use of resources in science. Publishers remain focused almost exclusively on positive, novel findings, for obvious reasons, one of them being that researchers share this focus. We can stimulate change by showing that we value our negative results and are willing to try to publish them, whether in a new journal (as Priyanka Srivastava suggests) or through some of the following existing channels.

An increasing number of journals like PLOS ONE, Scientific Reports, and BMC Research Notes encourage submission of negative results. Submit your work to them, and if there's unexpected resistance based on novelty, don't hesitate to remind the editor and reviewers that negative findings are welcome.

Even high-impact journals are often willing to consider negative results, especially as letters to the editor. In one of my research areas, Nature Biotechnology recently published a negative results paper. Make a strong case in the submission letter accompanying your manuscript, and you might be surprised.

Above, Christophe Pampoulie helpfully linked to existing negative results journals. These journals will be successful only to the extent they receive high-quality submissions, so consider supporting them.

Although peer-reviewed publication should be the ultimate goal, there are ways to share your negative results with the community prior to publication. For example, you can : submit a preprint to http://biorxiv.org/ , which encourages submission of "contradictory" findings; share your data through public, curated archives; comment on your new findings (ideally linking to the data) on PubMed Commons http://www.ncbi.nlm.nih.gov/pubmedcommons/ ; or post on your own website.

The idea is good. Publishing negative results can let the science community know your negative results so that those who are interested in the same question do not have to waste their effort. However, it is difficult to judge how reliable "the negative results" are. There are so many factors that can lead to the negative results, such as the rationale of designing experiments, the appropriate techniques applied, the doses and time allowed to detect the effects, etc. More importantly the subtle effects, often interpreted as negative results, are difficult to be noticed, and those subtle effects could be important. 

We (Journal of Negative Results in Ecology and Evolutionary Biology - http://jnr-eeb.org/index.php/jnr) are always looking for submissions, including articles aimed towards understanding better what "negative results" are or articles that may argue against what we are trying to accomplish. How often do accepted methodologies fail, where the results are not published and the next person comes along, spends a few years, funding, effort, trying the same thing, only to fail again? How many researchers selectively ignore parts of their results because they don't support some current dogma? Publishing work where these items are addressed can help build to foundation for future work. It's not meant to be an outlet for poor research. Interestingly, J of Negative Results in biomedical science really took off although their journal started after ours. This does hint at the fact that when lives are at risk, people seem to understand the purpose of and need for publishing negative results.

f1000research (http://f1000research.com/) publishes papers with negative results.

I think this is a great ideas.  However, we have to have a clear definition of what are the negative results that could be publishable. 

Negative results are really important part of our scientific knowledge, however, they are required to meet several criteria to be really valuable. Firstly, statistical testing should be much more rigorous than it is usually met in standard papers. It comes from the fact that negative results are often a product of small sample size. Therefore, papers/journals reporting negative results should focus on robust statistical analysis, namely power tests. If statistical tests are weak (due to small sample size) then you have fairly high chance to get negative results (so, increasing chances of the second type of error in the testing of hypotheses). Hence, it is clear, that scientists would accept negative results only if data analysis is thorough and much more advanced than in case of statistically significant results. Secondly, from my own experience I know that all experiments have many limitations. Actually, it is impossible to perform an excellent experiment in science. One of my colleagues said once: "give me any experimental design, I would find at least three weak points in it". I have to admit this statement seems true, because performing any experiment is always some kind of simplification of the reality. Thirdly, one has to think deeply about the tested hypotheses themselves. Are they really correct? Do former knowledge allows us to state them? To wing things up... negative results are always (I have got them a lot) a good “pit-stop” in our work, to analyze all our work, to review current knowledge and find possible explanations (new theories, shortcomings in the design etc.).

Best luck and only significant results in 2015

Piotr

The question of Priyanka Srivastava is really good and follow up discussion seems very interesting. Dr. Subhash Chandra gave nice definition of negative results. In my opinion, choosing specialized negative research journals to publish negative research findings would be negative approach. If one gets new findings those contradict to the exiting knowledge on the subject by following systematic scientific approach and methodologies then I would treat the results as ‘positive or corrective’ instead of negative. In this situation, the forum to publish such research findings should be well established journals so that the scientific community should have access to the new update. I would say this is how we progressed till date through developments in history of science. Therefore, I agree with the idea of Dr. Kenneth W Witwer to publish the novel and reproducible research findings either in well-known journals or community prior to publication. But, I would be more in support of established journals.

For example, recently we published an article in the journal Current Microbiology, where we found that microbial contaminants in the original cultures deposited in microbial cell culture banks can be mistreated as original cultures and their use can give bad/good name to original cultures in terms of specific characteristics although they may not have those inherent abilities.

http://www.ncbi.nlm.nih.gov/pubmed/25085545

These kinds of things of contamination have happened in case of mammalian cell lines also, where researchers showed contamination and replacement of original cells by HeLa cells, which were treated as original cell types. These findings were published in well-established journals like Nature. These results changed the course of the cell culture discipline.

I think, whether research results should be termed as positive or negative will also depend on the type of research area or experiment.

Great Idea !

Hi,

My opinion is that the negative results has more values than the positive one. Because negative results are the key to find a positive results. "Nonsense is better than Nothing" 

I fully agree with Subhash Chandra and others stating that negative results are really important for research progress. 

First you define your research question. Then you put forward a hypothesis that can be experimentally tested. Your experiments should always be conducted in a reproducible manner and be scientifically (i. e., choosing the right method(s) to test your hypothesis) and technically (e. g., including controls, replicates, clear recording and documentation of data, etc.) sound. If the hypothesis you put forward is true and your experimental data shows that then you might say that your results are positive. If not, they might be classified as negative. But the results (positive or negative) are wrong if you made some errors along the way of addressing your research question.

The general problem is that for many researchers a negative but correct result is disappointing because they expected their original idea to be true. These feelings are human but they should have no place in rational science. 

I appreciate all the suggestion. Every negative result has their own positive value if we are aware of advantages/applications and disadvantages of the result we got. Protocol has to be correct, but protocol varies from person to person coz we standardize protocol by our own way.  

What other factor which will affect this is "Negative Results". Most of the researcher, from the name of the journal -Negative results won't look at it coz we are getting so specific for our research and results. Other factor is "Impact factor".  What you think.? 

I would like to say, papers should describe the title, what they were trying to get, what other milestone they got as negative result, protocols they followed and applications/advantages as well as disadvantages of the results they got. That article should be open for all discussion.

I am an Associate Editor at the Journal of Negative Results - Ecology and Evolutionary Biology (http://www.jnr-eeb.org/), so I can tell you a little about our experience.

This journal was set up in 2004, with many of the aims highlighted above (appreciating and publicising the importance of Negative Results (roughly defined as a statistical result that does not reject the null hypothesis being tested in an appropriately designed experiment), e.g., avoiding unnecessary replication of experiments that do not produce significant results, highlighting the importance of considering experimental statistical power, etc).

All the editors involved with this journal remain positive about the importance of negative results (see what I did there?), however, this enthusiasm is not filtering through to our academic community as widely as we'd hoped.

We've been publishing for 10 years now, but still receive very few submissions at JNR-EEB. We can only speculate on the reasons for this (pressure to publish in more high-impact journals, loss of researcher focus following failure to generate a result that supports their pet theory, the rise of the mega-journal that focuses only on methodological rigour...), though in many ways, we don't want to have a separate journal for negative results - they should be published in the same journals as the 'positive' results. However, I doubt this is really happening - the rate of publication of p > 0.05 results is almost certainly lower than the rate at which they arise following experimental/observational work.

It remains an important topic, so thanks for bringing up this discussion!

Hello Mike, thankyou for appreciating this topic. As a researcher, if I would look towards journal of negative result, I won't pay interest and move forward to look towards some good impact factor articles. This is a general philosophy. Most of the researchers are not even aware, that few journal exist that publish negative results. Now a days, researchers are getting more specific towards their expected results, that is why they won't pay attention towards negative results. 

What my point is: Journal should be open for discussion, just like research gate where we can discuss protocols and take ideas, generate ideas. Negative results should be discussed briefly then only we can realize their importance, we can add up some new methods, can find out what we are encountering and explore more about that point and that is called science.. Right.. :) 

Hello Marco.. Yes..! That's what..! Results are results..! 

I hope you don't mind another shameless plug for http://www.jnr-eeb.org/.

We are completely open access and do not have any Article Processing Charges. Publishing in JNR-EEB is (financially) free for everyone!

We actually received a submission last week following this discussion - it may have been coincidental, but hopefully interested people are reading and considering submitting their work to us.

That is great Mike. 

See this for more on negative results 

http://www.tc.umn.edu/~allch001/papers/neg-res.pdf

if the experimental design and statistical analyses are sound and have no mistakes, it is possible to publish even if the results were negative. you can, for example, refer to author instruction os "scientific reports" journals published by Nature group as an open access journal. they mention that even negative results of correctly conducted experiments will be accepted for publication. I personally believe that negative results can be more important t if the experimental design and data analysis were correct. Most of the extraordinary findings come from what thought to be negative results. Especially in biological sciences, having your results exactly as you've expected is meaningless; why did you conduct the experiment, if you will reject results away from your expectations?  

Sound experimental design, data analyses and honest observations from repeated experiments will never end in meaningless results.