First thing, I would present this to the external lab. Perhaps they can look back at notes, strip charts or other outputs to check for mistake. Then if that fails, present the truth with the possible erroneous value and cross checked with other methods, and discuss briefly in findings. Like suggested, if the whole journal paper has no value with this issue included, then you will probably get some bad marks from editor or reviewers, possible suggestions to recover. When you say the value from lab is extremely unlikely or unnatural, it is likely the lab just made a mistake, it happens. It would be better for them to review the circumstance and your cross checking, and agree there was a mistake and then use your value based on their agreement, with a footnote perhaps briefly mentioning this attached to the value. Other issues can develop such as if your cross checking followed much later in time and samples were not preserved and/or stored properly. You may have to justify your work too.

Dear Thomas,

If the unreliable or insignificant results are just a part of the set of findings, then a journal will likely accept the research article. However, if the paper is based only on the unexpected, unreliable, or insignificant results, the journals won't come closer. The discussion section is a better place to incorporate interpretations of unexpected results. Somethings should be mentioned in the results section, too.

Best

Ibrahim

First thing, I would present this to the external lab. Perhaps they can look back at notes, strip charts or other outputs to check for mistake. Then if that fails, present the truth with the possible erroneous value and cross checked with other methods, and discuss briefly in findings. Like suggested, if the whole journal paper has no value with this issue included, then you will probably get some bad marks from editor or reviewers, possible suggestions to recover. When you say the value from lab is extremely unlikely or unnatural, it is likely the lab just made a mistake, it happens. It would be better for them to review the circumstance and your cross checking, and agree there was a mistake and then use your value based on their agreement, with a footnote perhaps briefly mentioning this attached to the value. Other issues can develop such as if your cross checking followed much later in time and samples were not preserved and/or stored properly. You may have to justify your work too.

Once a research article is published, readers take it for granted the authenticity of not only the data but even the inferences drawn out. Therefore, author of a research article has unwritten responsibility to communicate the right information. There are a couple of tests you can perform before accepting any data - for example, water samples - there is a strong correlation between the pH and the other parameters which need be cross checked. Ionic balance is once such exercise you could lean on.

If the results are unreliable and you know it well, better stay away from publishing ...its my personal opinion. .

If your data are able to be published then they are not unreliable, so don't even mention it. It's an oxymoron! :)

Greetings

If you are sure of the unreliable data, for sure the first step is trying to repeat the analysis to get the data you need and feel confident with. Unreliable data, will result in unreliable interpretation and conclusions (although other people don't know it,.... but most importantly, you know it as a responsible scientist).

If it is not possible to repeat and obtain reliable data, you should definitely omit this part and do not bother yourself and waste your time interpreting it. Your results and conclusions may affect the life of someone in the world (either physically, financially, socially, etc.).

Good luck

Dear Mr. Fungenzi,

every paper should be based on reliable results obtained by means of trustworthy methods. Your paper, I guess, is not intended to present the back and forth of methodological processes but aimed at providing a solution to, in your case, a pedological issue. As papers used to be written concise and so as to get the point straight to the heart any deviation from the topic in the discussion directing your thoughts to methods and the use of external and internal labs should be avoided. Be focused and punchy in the presentation of your scientific results and keep off roaming around. This statement is based on practical experience gathered as author, reviewer and editor.

H.G.Dill

Why to publish unreliable results? I would suggest even if unreliable results have been published somehow by someone, as scientists it is our duty to point out the unreliability of such results.

Hay Thomas Fungenzi,

I would like to mention here as you aren't sure on similar method used at triple cross checking that of external lab do. At this moment, I could suggest to re-analyze same parameters in that external lab, if possible. And confirm its reliability or unreliability, either then for your publication.

Just adding more, Lab to Lab's results may vary. If, there is substantial difference between your reliable (expected) and unreliable (received) results, I guess something must have missed while analyzing samples at labs end.

( The problem is, that the method I used is not the same as the initial method (unavailable at my lab), but is supposed to measure the same variable)