Sometimes I like to write about history- i.e. things I’ve learned that you might be able to benefit from. Other times, I like to gather your collective opinions about one topic or another because it helps me firm up my own. This is going to be one of those posts.
I always thought I had pretty firmly set up in my head the guidelines that I follow for including someone as an author on a paper. Lately, circumstances have forced me to confront (again!) the fact that different people have different, and sometimes vastly, vastly different ideas of what contributions constitute criteria for authorship on a scientific paper. This has been an uncomfortable examination of my own standards and those of others.
Let’s just start with my own view. I include members of my lab on papers to which they have directly contributed data, and/or made an intellectual contribution to the work in the manuscript. Generation of a reagent only counts in a particular set of circumstances- obviously just making the TAE isn’t enough- but building a particular set of strains (that are not published elsewhere) that were used in a figure or two, would be a sufficient contribution for authorship. Seems pretty straight forward so far, right?
Here is where it gets complicated.
Scenario #1
Let’s say that you have 3 employees- doesn’t matter if they are students, postdocs, techs or whatever, working on a single project, basically using the same technique to get at the same question. (And we almost certainly discuss whether this is a great strategy for giving people projects to begin with, and for the record I don’t do this in my lab…but that is not the question right now.) Scientist A and Scientist B do lots of experiments, but the data is never great enough to appear in print. Scientist C has awesome hands, and gets the thing to work- and work beyond your wildest imagination. Do you include all of them on the paper that eventually makes its way out of the lab on this work, or not? Does it matter if they are in the lab concurrently or not?
Now let’s imagine that Scientist B, despite the fact that his/her experiments were a total bust, was letting Scientist C look over his/her notebook- and they were spending a lot of time troubleshooting together over beers on Friday afternoon-…. does that change your opinion of who should be on the paper…?
Scenario #2
We are going to build on scenario #1 up there for this. For your own lab members you decide – in the case where Scientist C was wildly successful without input from Scientist B (i.e. both A and B performed failed experiments)- you decide that only Scientist C should be on the paper. But you’ve got a collaborator who has to make similar sorts of decisions to determine who will be authors on a collaborative paper… from both of your labs. Your collaborator has the philosophy that the lab is a team, and that no matter whether or not particular individuals produced actual data that appears in the paper….
I think you see where I am going here- how do you reconcile who goes on the paper from each lab. Does each PI decide independently who from their labs should be included, or do the PIs have to have some sort of agreement about what constitutes enough of a contribution to be included as an author?
I gave two pretty simple examples, but I know (and all of you probably do as well) that it is almost never as simple as this. There are careers to consider- and productivity is a major part of who advances to the next level. I have colleagues that put the whole lab on the paper, because then everyone has a longer CV. I was talking with a colleague last night- and after listening to the stories he/she told- I came away from feeling like authorship disagreements and disputes are more the rule than the exception.
So, what do y’all think??
Blue Lab Coats 
I hate being scientist B (who shared the techniques over beers).
“I have colleagues that put the whole lab on the paper, because then everyone has a longer CV”
ARGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGG
I *really* hate this shit. It’s a total dick swinging move. It’s not about the paper. It’s about “look how many people work for me!$$!$! look at my connections!$! I swing my dick and it hits everyone!!!$$! mwa ha ha I rule the WERLDDD!” The good ol boys love to pad each others, their buddies, their buddies’ dogs, and the tree outside’s CVs. Ask any of them about the specifics and they refer you to buddy moron 2, then 3,… 54. Now, there are papers where there *really* are shitloads of people who contributed to the actual project, either in writing or gathering data or analyzing results (human genome papers). But there’s a big difference between throw the baby and the bathwater and all the toys in sight onto the paper vs. everyone who contributed to part of the project on the paper. Some of the zillion author papers in my field are totally laughable because they are IF=1. *eyeroll* It’s just CV padding, but to a zonked search committee and tenure review, these padding-happy morons appear well-connected and productive.
/rant
alrighty then….
Scenario 1: C has output, C gets authorship. A and B have nothing to show, then there’s nothing to show. A and B go to Ack.
Scenario 2: if your collaborator is first author and s/he decides that since A, B, and C were all involved in the project and that’s the way s/he sets authorship up, then A, B, and C get authorship. if you are first author, then you decide (in this case, C would get authorship).
Glad you’re back! hope you had a nice trip.
Case 1: C. Case 1A: yes, I change, and B gets on there also. If I’m B – and I have been – then as soon as it get to the point where I think I’m doing enough substantive input to this paper that it merits an authorship, I mention that (verbally, and follow-up in email) just to avoid unpleasantness later.
Case 2 sucks. I had a recentish paper where I was expecting to be maybe author 3 of 5 (big paper, I contributed 2 figures and all of the experimental work/writing/analysis that went with them). Instead I ended up 4th of 20 or so because of all the ‘me too’ folk. Annoying; but only relatively minor suckage, all things considered.
I am going to have to face these Qs soon as PI; and I think I agree that the first author (and you had better hope *that* is clear!) gets the deciding vote.
I say A+B+C get authorship in both cases, since they all did work related to the manuscript. *If* A+B did no related work to the manuscript, but simply belong to the lab, then they should not be included (i.e. no padding). I’ve had plenty of “who the f*** are these people” go on my first author papers … and I’m sick of it.
I am not a PI, but I’m somewhat concerned that A & B failed to get results.
If A & B got results but they weren’t significant or had too much noise, but C generated a data set with crisp results, then all sets of data should go into analysis unless there are good reasons to exclude some.
I think these questions matter for a more complex look into authorship issues, as well as a look into the validity of the data.
Re #1: If scientists A and B had input into the design and analysis of the study, even though their own attempts to produce data were unsuccessful, they should be included as co-authors.
Re #2: I’m with jc on this one – I hate the “entire lab on each paper” bullshit as it pads cvs unnecessarily at the expense of those of us that actually had to work our asses of for every single one of our publications. If I was collaborating with another PI and one person from his/her lab did the hands-on work, I would be loathe to include the other gazillion lab members on the paper just because that was the done thing in Collaborator’s lab.
Thus endeth my 2 cents worth.
In relation to the notion that it is the first author who decides authorship issues, this is laughably wrong, at least in the biomedical sciences. It is the senior author from each lab who decides who from her lab is an author on the paper, and it is the senior authors collectively who negotiate author order.
It is almost always best to discuss these things before beginning projects, either within one’s own lab or in multi-lab collaborations. For example, I recently received an e-mail from a colleague proposing a collaboration between our labs.
She listed the various experimental action items that were to comprise the collaboration, including which were to occur in her lab and which in mine. She then stated that, since it appeared to her that the most-difficult, time-consuming, and important experiments were going to occur in my lab, that my trainees who performs those experiments ought to be first author and I ought to be senior author.
I’m with PiT, that I’d be loath to throw A/B off the authorship train for failure to contribute data if it is the case that they significantly advanced the intellectual rationale or contributed over half of the experimental troubleshooting. That stuff matters.
As PI, I would talk with A/B (particularly B, it appears) about what she perceives as her authorship claim. Even though people don’t want to pass up “free” authorship, most reasonable scientists will take seriously the question of whether or not their contribution reached the authorship benchmark, and it’s a great mentoring opportunity too–to lay out the case pro and con and invite B to think about it carefully.
As for question 2, ugh. The problem is that adding more names from your lab doesn’t actually help C, who is then simply saddled with even more co-authors; and the marginal value to A or B of being Nth author on a 10-author paper is not that high. In this situation I would push very hard for an “author contribution” statement, whichever way you decide about A/B authorship. I know some people (hi, CPP!) say those are crap because no one reads them, but I think they provide a certain sense of justice for all involved.
It makes a great deal of issue here whether A, B, C are graduate-students/postdocs or technicians. (If they are technicians, then they are getting paid to do the technique. Providing an authorship is a reward for work above and beyond running gels and slicing brains.) But if they are GS/PD, then the problem here is that scenario #1 presents a deeper problem than authorship. Instead, it presents a failure of mentoring by the PI. Why were A and B not able to contribute to the paper? Presumably the work by these graduate students/postdocs is not just running some straightforward tests, it includes intellectual contribution to the planning, the writing, and the science. At which point, the PI needs to be on top of why A and B are not contributing to the project long before the final paper is submitted.
if A and B contributed significantly to working the technique out, and C was able to make it work due to all of the groundwork laid by A and B, they should get recognition for their work. if they did nothing but fuck up, that’s another story.
how about a situation where A from collaborator B’s lab teaches C how to use a piece of equipment for about 20 minutes, and C puts in 1000+ hours on said equipment generating totally amazingly beautiful data. meanwhile, A is processing all the data that C generates behind C’s back before C has an opportunity to even look at it. A has no fucking business processing C’s data, and had no hand in the experimental design or concept or anything. this generates massive controversy in the lab when C calls bullshit on A for his general asshattery. but A has spent a lot of time processing data, and did a fairly good job of it. C needs to learn how to process this type of data, so C redoes the entire thing herself and uses her results (which do match A’s). does A get authorship?
there is no reason for this type of shit to occur, right? someone tell me this is bullshit.
Qaz- I know you are right to some extent about whether or not it matters whether the people being considered are postdocs/techs/students- but I wanted to leave that out for the moment and focus on the immediate question. Which itself, isn’t black and white.
C PP- I am in total agreement with you that it is the SENIOR author or authors who decide who is on the paper. I also think it is a great idea to at least talk about this up front- but it is not always so simple. You and I both know that during the course of a project things can change-people who weren’t thought up front to be in line to make a major contribution can be the ones who break the thing open in the end. So in my view there needs to be a re-negotiation or at least room for discussion if something like this should occur.
jc- I’m with you all the way in the everyone and the kitchen sink department. Thanks for the welcome back- I’ve been around for a week or so but was just swamped. 🙂
Dr. J and Mrs. H- Let’s say, just for the sake of arguement that A/B’s data was unusable, and C – on his/her own- had the insight that a new technique was necessary. Now what?
Sara- There are lots of scenarios in genetics in which I can envision this happening with A/B that don’t make Cs data invalid.
Becca- I’ve been scientist B a whole bunch of times, and it’s awesome to see someone take an idea you contributed and run with it… but it feels less awesome when your own experiments aren’t working… I’ll give you that!
I’m sure I’m on the strict end of things, but I think every author should be able to extemporaneously explain every figure (not necessarily details about which antibody was used, but what the figures mean) . They should be able to have a conversation with another scientist about the paper their name is on. If that means fewer technicians who did a lot of grunt work without knowing why they did it, so be it — better to have the techs who do know what they are doing on a 5 person paper instead of a 10 person paper.
A PI/postdoc/student should think about authorship on a paper in terms of retractions — if your collaborator from another lab came to you and said that the paper needed to be retracted, would your defense to your department chair be “well, I didn’t really have much to do with that work” — if so, you probably shouldn’t be on the paper in the first place. If you don’t feel like you intellectually own the work, then you shouldn’t be on the paper.
Absolutely!
I just took the CITI certification course on the responsible conduct of research for my IRB stuff, and actually this question is not a matter of the opinion of just the senior author (or any of the authors). There are some pretty strict ethical guidelines about who is and is not eligible for authorship, and if people either do or do not include authors who do not fit in those guidelines it can be grounds for investigation into research misconduct.
I even had to retake a couple portions of the quizzes on this stuff because some of it is a lot stricter than I ever thought it would be. Here’s what they said in the course:
1. Any person who contributed to the intellectual content of the manuscript or who had control over final editing and approval of the manuscript for submission should be an author. Anyone who did not participate in those processes should not. According to the CITI course, people who generated data but did not contribute to their interpretation/analysis are kind of a fuzzy line: the leader of the project should decide if they contributed “substantially” enough or not. But anyone who contributed key intellectual parts (whether by creating data, troubleshooting data/problem solving, or interpreting someone else’s data) needs to be included.
2. Authorship should be discussed at the beginning of a project, but should only be FINALLY DECIDED once the work is finished, the data interpreted and the manuscript is in final or nearly final form.
leigh: I think with regards to my #1 up there, in the situation you describe the person who surreptitiously analyzed someone else’s data still SHOULD NOT be an author because they were unethical to ‘steal’ that data without permission/knowledge of the project leader/data generator in the first place.
hm, I realise that I am harsher than I thought I would be. If it is partly because I have been excluded from papers or if it is from the stand point “it isn’t fair” I don’t know.
In short, in the scenarios you paint I am not sure that A should have authorship at all. If A didn’t contribute intellectually nor with data (and it sucks to be excluded because YOU couldn’t make assays work that others did but if you compare “making strains” then it doesn’t pay off if you didn¨t succeed making them, then you will not get on the papaer either).
The whole fair thing gets more complicated if you think about the “other lab” as well. Being the second lab and then all of a sudden get 5 people on the paper because people “tried” sucks. Sorry to say but that isn’t fair either.
I would go with, people get on the paper if they contributed data in a figure or background data the work built on (so called supplement data in some cases) or if they trouble shot / discussed set up, procedures and research experiments more than what would be expected in a lab. (Guess this is where I make it hard on myself, I think that a its expected of people in a lab to talk and communicate research without assuming authorship etc. It’s part of science.
but I am not a PI so, I am sure it will be harder when you are up there …..
Put everyone who was involved with the work on the paper. First authors worried that middle authors take away some of their credit are wrong. “Ethical guidelines” some faceless committee invented arbitrarily are to be ignored at the discetion of the paper’s last author.
1) In the first scenario, regardless of how much or how little scientist C talked to anybody else, I want to know why C’s results were so much better than anybody else’s. This had better be abundantly clear before C’s data goes to print, regardless of the author list. I don’t want to find anything out the hard way later on.
2) If scientists A and B worked together, and either one of them communicates with C in a productive manner, then the work of both is shaping the intellectual approach. I think both of them deserve authorship.
3) To tie the first two points together, if there’s a reason why C got it to work where A and B failed, and this reason tells us something about what’s going on. For instance, if A and B failed to control variable X with sufficient precision, while C was more careful with it, and that’s what made the difference, then comparing the work of AB vs. C tells us that variable X is really important here. That should probably go into the paper, further justifying their inclusion in the author list. This presumes, of course, that their error wasn’t just simple sloppiness. But if they failed to control some variable because nobody realized that it might matter, then their data is important to demonstrate that C’s control of that variable is the key.
whimple- Oh how I’ve missed you. But on to your point- some journals these days have a section at the end of the manuscript where author contributions are clearly delineated… in the case of A and B’s experiments not working, and C coming along trying something new and getting it to work… you include A and B on the author list- but how do you delineate their contribution to the paper?
And Alex, let’s just say for the sake of discussion A and B were using a particular technique, C came along with a strong bias that that technique wouldn’t work – and tried a new technique.
chall- I’m harsher than you, believe me.
Whimple: ““Ethical guidelines” some faceless committee invented arbitrarily are to be ignored at the discetion of the paper’s last author.”
You can say that all you want, but the NIH and NSF subscribe to the guidelines taught in the CITI course and all it takes is a diligent whistle blower to put your name on the website where they publish the names of people found to violate responsible conduct of research and good luck getting funded again.
Well, from my perspective, it’s completely retarded and unfair for the PI to have pitted three researchers against each other in the first place. With that in mind (given that fairness has already been thrown out the window), I can’t see the harm in including authors A and B on the manuscript. Few people give a rat’s ass about 2nd/3rd authorship anyhow (in the context of a CV) unless it’s published in Nature or Cell or something. It’s one thing to talk about authorship “rules”, and another thing when we realize that these are real people involved. Real people who have sweated their science sweat only to be essentially scooped within their own lab.
As for the cross-lab collaborative issue, I don’t even know. Sounds too complicated to get the PIs to agree, but sounds unfair to have different standards for different labs.
For any potential junior author X, if they go to a conference and have lunch with a future employer/ collaborator, can they intelligently explain their contribution?
Author statements don’t hurt. If lab A lists one person as having done task Q, and lab B required 5 people to do an equivalent task, then that sort of speaks for itself.
It can get sticky if a number of people are working on similar experiments utilizing essentially the same technique. I cannot imagine that, being in the same lab, meeting regularly and discussing among each other, they will not provide each other with valuable input. I mean, I give my students very different projects but we have regular group meetings and I encourage them to discuss their research projects among themselves — and on occasion I have felt that some would have deserved> authorship even though it was not their main project.
However, I thought that in the biomedical sciences this would be less of a problem, given the emphasis ascribed to being first author. I would imagine that there would be no problem putting all of the students/postdocs who have conceivably contributed intellectually and in practical terms, while ensuring first authorship for the person for whom that is the “main” project.
Pingback: ScienceBlogs Channel : Life Science | BlogCABLE.COM
I’m not sure there’s a ‘should’ or ‘should not’ in this circumstance. It all depends on the lab atmosphere the PI wants to foster. If the PI wants to promote competition then let the best/fastest results win. If the PI wants collaboration within the lab then everyone who produces work that contributes toward the project should be on the paper.
Of course, if someone wants to foster a collaborative atmosphere, they’re probably not going to put 3 scientists on the same project and same experiment.
Another vote for Sara + Alex. If C invented or used a new technique, then C needs to validate that technique. Several different ways, for preference. Invention of a new technique can be a separate paper or patent, actually, if it’s that awesome and so sensitive. I would like to think that A and B weren’t using half-assed techniques in the first place, I would hope that they would use the gold standard for the field rather than wasting time and reagents. In which case, C has some explaining to do. I have seen a lot of Scenario #1s, and I gotta tell you, the overwhelming majority of them DO tend towards Journal of Irreproducible Results sort of thing. As soon as you say “beyond your wildest dreams,” it’s time to be suspicious. Very very VERY suspicious. Demand all raw data, untouched, unlabeled, un-‘Shopped, including notebooks and random scribbles on paper towels. If for ANY reason it cannot be obtained, or is penciled instead of inked, do not publish. It may be that C is honest as the day is long on June 20th in Alaska, but you should suddenly take a deep and abiding interest in C’s documentation. You know, to find out why it works and what protocols C is using. And you need to demand something like n=5, because while one really great blot might be publication quality, there better be three or four other ones that confirm it was not just C’s lucky day.
In Scenario 2, I would be the collaborator telling you that you are screwing over A and B and making hard feelings. In our lab, it’s pretty rare for any major experiment (i.e. one expected to result in a paper) to get done without prior discussion with colleagues. We do share responsibilities for construct design, cloning, expression analysis etc. and for the most part, the person to whom the project is assigned can have a reasonable expectation of being first author. But since you asked about PIs, I would say that both PIs need to agree beforehand what constitutes a contribution.
Perhaps there is a little too much being read into drdrA’s original scenario regarding the hows and why Scientist C’s results were better than the others. As one who spent months trying to reproduce a published biochem assay during grad school, I can attest to the fact that these things can often take a lot of troubleshooting or sometimes just a fresh pair of eyes (there were 3 of us working on it at one point – the other 2 gave up in despair). I’ve also worked on analyzing a protein that few labs had successfully measured using western blotting (hard to believe, huh!?) and that none had shown in our organ of interest using immunohistochemistry (we were told it was “impossible”) – a lot of blood, sweat, tears and troubleshooting finally yielded success in both where others in my lab had failed.
And yes, in all of the above, the appropriate positive/negative controls, standards, etc proved that my techniques produced valid data. For the latter example, we (my postdoc lab and I) are going to be the first to publish immuno images for this protein in our organ of choice and the data we got as a result have been the jumping-off point for new work in my postdoc mentor’s lab.
Just because 2 people couldn’t do something that doesn’t mean it’s impossible!
Everyone- Let’s assume that there was no scientific misconduct on the part of C, or anyone else. We can talk about scientific misconduct- but that’s going to take a whole different post! And Writedit and Drugmonkey do this much better than I could ever imagine to!
Putting aside all the “this scenario should never happen if you’re a good PI” stuff, which is silly because it happens no matter how diligent the PI is… BLC is not at liberty to give more details so let’s give her the benefit of the doubt. Heck, this exact scenario happens all the time after A and B have left the lab or moved on to other projects. A could have been a former post-doc who tried something new and it never went anywhere, B a rotation student who picked it up and got one or two interesting data points but nothing usable and dropped the project but maintained interest in its outcome, and C a dedicated grad student who finally had an insight into how to make it work… you get the idea.
Anyway, to add to Dr.J.Mrs.H.’s suggestion to gauge B’s feeling towards B’s claim of authorship, I’d gauge C’s feelings towards A and B’s contributions. If C is at all honest, C will give a sense whether C feels that C’s contributions were made completely independently of A and B’s (note I’m not being naive — if C is not being honest it will often be clear to the PI after talking to A and B anyway). In this scenario, it already seems highly likely that C did benefit from B’s input and experience, so unless C has a research-based objection (ie B was urging me to try it B’s way and I persevered and got my own way to work) it seems clear B can go on the paper at the PI’s discretion. I agree with CPP that middle authorship is not winning anyone undeserved grants or tenure-track positions, nor does it dilute C’s accomplishments. If C also benefitted from A’s experience, as judged by the PI after talking to A, B, and C, then A can also go on.
IMHO, an author contribution statement is a must in this situation, in which it is stated that B (or A and B) were involved in “preliminary experimental design for the fly mating procedure of the red-eyed YFP2 deletion strain”, or some other disclaimer. For me, the key to the disclaimer is to make it very specific and include a qualifier like “preliminary” to show that the contribution was only on one set of experiments, and did not lead directly to the data in the figure.
MBench-
Heck, this exact scenario happens all the time after A and B have left the lab or moved on to other projects. A could have been a former post-doc who tried something new and it never went anywhere, B a rotation student who picked it up and got one or two interesting data points but nothing usable and dropped the project but maintained interest in its outcome, and C a dedicated grad student who finally had an insight into how to make it work… you get the idea.
I have wanted to say exactly this- I venture to say that this is not that rare- that someone (let’s just say A) doesn’t get a project to go, and moves on…. the new guy/or gal (C)… comes along and gets the thing to work… NOW WHAT. To me, A is not an author in that case.
BLC — Definitely. If A made no progress and has moved on either literally or figuratively, to me that’s a no-brainer. I have a feeling that there might be less gray area in the scientific sense than we imagine (at least for A), but the complications come when someone regards something as “MY project.” As in “how come C published something on MY project?” Um, A, perhaps because you abandoned it years ago?
Hi everyone!
Sharing my two cents…
Scenario #1.
a) In my opinion, only C should be included in the paper as only his data appears in the MS.
b) (this is when you suggested the B talked with C). In this scenario, it looks like B had an intellectual participation in the generation of the data, or as you put it
“made an intellectual contribution to the work in the manuscript”. He worked and also suggested changes to be made to the protocol and helped in the troubleshooting which ultimately led to data generation (by C). In this case, B can be included in the byline.
Scenario #2
Putting al lab members in a MS (just for the sake of it) it’s just unjustifiable crap.
I guess I’d take the path where I talk with the other PI regarding what constitutes enough contribution to be included as an author. However, as some has stated, I would definitely talk about this at the beginning of the project. This is a lot simpler than talking about “who” will be included (the sort of discussion that typically arises after a project), as here you’ll be discussing the conditions, not the actual individuals.
I’ll also strongly push for a strict “author contribution” statement.
Cheers!
I would definitely talk about this at the beginning of the project.
In practice, this mostly turns out to be a waste of time. There usually aren’t author disputes, and when there are, they usually arise because the project went in directions that were not predictable at the start of the project.
As I mentioned, you should not discuss the individuals, but rather the merit to be counted in as a author. This can (and luckily has been successfully applied in my lab) be done at the beginning.
But you are certainly right in the context you were discussing.
Cheers,
I have a real life situation and I need some advice. I collaborated with Lab 1 for many years. They sent me really nice samples and we tried to analyze them. I couldn’t quite get the analysis to work but they kept sending more samples as we improved our analysis protocols. About the time we were really starting to nail down the analysis protocols we started a collaboration with a second lab, call them Lab 2. Lab 2 provided samples that in many respects were identical to those provided earlier by Lab 1. We finally got our analysis to work perfectly using the samples from Lab 2. We got excited and wrote a paper with Lab 2 describing the results. We kind of forgot to include Lab 1 in the writing and none of the figures in the paper come from samples provided by Lab 1. So I’m thinking that Lab 1 should not be included as authors. What do you guys think?
Uf, that’s a tough one. However, I wonder… how much was Lab 1 involved in the generation, troubleshooting,etc of the methods to analyze the samples?
I’m just thinking that if they were involved in the generation of this “analysis method”, which you later used to analyze data from another lab, then they should be included, as you wouldn’t have been able to analyze data from lab 2 without the intellectual and/or physical input by lab 1. This is regardless of them not having their data included in the final manuscript.
However, if they (lab1) just gave you the samples for you to “figure out how to analyze them”, and then Lab2 came along and did the same, then is your method that lab1 and lab2 are using, and they should both include you, independent of each other.
If the samples used in the paper came from Lab 2 and the writing and figures all came from your lab and Lab 2, it’s hard to justify including Lab 1 authors on the paper unless they had a clear and demonstrable intellectual contribution. Although that may seem unfortunate for Lab 1, that’s just part of how projects develop in science. I believe in being generous with authorship, but it doesn’t sound like a good case could be made here for including Lab 1 authors, especially since their samples did not work in your analysis for whatever reason. From a different perspective, if I were in Lab 1, I’d feel uncomfortable being included as an author on a paper to which I did not materially contribute.
In the case of authorship, it seems to me that trying to apply any kind of ethics metric only ends up making the issue even more complicated than it perhaps it really needs to be. Authorship is best handled with a flexibility based on common sense and diplomacy rather than by a rigid set of guidelines that one can guarantee will be hopelessly uninformative in a substantial proportion of any arising disputes.
The guidelines listed by Arlenna seem to support this view; being far from objective and clearly defined, they go no further than to state the obvious and leave the problem of exactly what metric should be used to gauge “contribution” up to the discretion of the PI. They certainly don’t resolve either of the scenarios depicted in this post as far as I can see.