Authorship arrangement...what seems fair?

[propsych]

Hi there, this is very long so I really appreciate any insights! so jumping straight into my problem:

I will be running a study this fall, for which my PI and I want to write a paper from. A colleague of mine has been helping a lot on the project preceding it (was needed to get parallel forms) and helping a bit on it itself with some editing and methods chats.

Below I outlined more specifically what happened, but basically I want to know if it’s fair that my colleague wants to be 3rd author on the manuscript and only write up let’s say a couple of paragraphs for the intro that pertain to the measures she added to the study (that she took from other labs)(her justification is she did a lot of the background work for the study before it). She wants us to add another RA as fourth author and have them do all of the future data analysis and write up of the results.

I believe this study would not be if it wasn’t for me, and the responsibility of all peices of these studies coming together and working were up to me. If I needed help with certain aspects I would delegate to RAs but I managed the project. My colleague did help a lot, however I sill feel that the main weight of it was on me.

Last fall I came to my PI about doing a study that was of more interest to me than what she was having me work on at the time. She agreed and I began developing the methodology. I then realized we didn’t have good enough dependent variables and so we created a normative study to create parallel forms for the actual study I wanted to run. Another memeber of this lab who is my senior, hopped onto the project and added some ideas for measures we could test and potentially use for the experimental study.

We ran the study and because this colleague is older and knows how to use spss, she managed data analysis this summer and is writing up the entire data analysis plan and results report.

We will now have parallel forms for the experimental study that will run this fall (the one I originally had the idea to run). I have been working on the irb (she has also been helping develop methods a bit/editing), trainings RAs, and basically getting the entire study organized and ready to go.

I am not sure if doing the work on the normative study justifies a third authorship on a future manuscript that uses the normative data, without doing additional work. Especially when the fourth author who would be doing all the results/analyses has also been working on data analysis this summer.


Am I right in feeling that this agreement is unfair, and if so, how should I approach this?

---

Members don't see this ad.

 

[Justanothergrad]

Authorship is based on contribution. If your question is 'does the person who helped select variables, ran your analysis because you weren't sure how to, and is writing some sections warrant inclusion' then my answer is yes. Why would they not? If someone wants to write a third of my papers and do the analysis I'll give them authorship. I take a rather inclusive view on authorship since it is better to be honest and equitable than look like you did it all (looking at authors who do metas on huge subjects with one author..). It also makes life easier.

What do you have to lose by including her since it's clearly ethical to do so? In the future, have conversations up front on the project start about expected contributions and product outcomes if you aren't sure about something.

 

[propsych]

Authorship is based on contribution. If your question is 'does the person who helped select variables, ran your analysis because you weren't sure how to, and is writing some sections warrant inclusion' then my answer is yes. Why would they not? If someone wants to write a third of my papers and do the analysis I'll give them authorship. I take a rather inclusive view on authorship since it is better to be honest and equitable than look like you did it all (looking at authors who do metas on huge subjects with one author..). It also makes life easier.

What do you have to lose by including her since it's clearly ethical to do so? In the future, have conversations up front on the project start about expected contributions and product outcomes if you aren't sure about something.

Yes I totally agree, I guess the specific thing I am asking is if it is fair if they don’t actually write much for the paper and have the 4th author do all of the analyses when it’s time for that.

 

[Justanothergrad]

Some professional advice,
1. They already did analyses and were part of the planning phase for the project that seems to have enabled your current project. I assume that paper wasn't published by itself (if it was this /may/ be different... but see point #2 and consider it several times). If you were uncomfortable with them being on this paper which utilizes their work, the time for that conversation has long sailed imho.

2. It doesn't hurt you to include someone who did something. It can hurt you to snub people. Be wise, not proud.

3. If you want more help, ask for it explicitly on specific tasks but don't abuse this and be willing to be the heavy lifter on your FA papers.

4. It's always wise to be able to do your own analyses.

5. Talk to your advisor

6. As for 'fair':

 

[propsych]

Some professional advice,
1. They already did analyses and were part of the planning phase for the project that seems to have enabled your current project. I assume that paper wasn't published by itself (if it was this /may/ be different... but see point #2 and consider it several times). If you were uncomfortable with them being on this paper which utilizes their work, the time for that conversation has long sailed imho.

2. It doesn't hurt you to include someone who did something. It can hurt you to snub people. Be wise, not proud.

3. If you want more help, ask for it explicitly on specific tasks but don't abuse this and be willing to be the heavy lifter on your FA papers.

4. It's always wise to be able to do your own analyses.

5. Talk to your advisor

6. As for 'fair':

Point taken haha! You’re right, I have a great working relationship with this person and she deserves to be on it (that was never the issue). I will ask explicitly if there is a particular aspect of the paper I would like her to consider taking on. I will be doing most of the writing as the second author on this paper, just thought a third author should contribute some writing as well considering we both did previous work on the enabling study so to speak.


That said, I go back to point #2 and relax

Thanks for the advice! Much appreciated

 

[WisNeuro]

If they are doing anything substantive on your project design, analysis, and/or manuscript, it would be unethical to leave them out of the authorship. Heck, when people use someone's database, that person often gets thrown on as a last author even without writing anything.

 

[Ollie123]

Agree with the above, I have no earthly idea why this person would NOT be an author as everything seems to point towards them being included. The key is substantive/intellectual contribution. Its extremely common for a statistician to develop the analysis plan and "outsource" the actual programming to someone junior (master's level statistician, student/post-doc). That junior person may or may not be included as an author. Just like the RA on the project who is just following orders doesn't necessarily get authorship.

Scientific writing is not like writing a novel. Authorship is not about "who put the most words on the page." The overwhelming majority of co-authors won't write a single word beyond some editing. Writing is often one of the less substantive parts of the process. If I had that level of involvement on a project and got cut from authorship, I would certainly never help that person again (and might be reporting them for ethical violations).

 

[propsych]

Agree with the above, I have no earthly idea why this person would NOT be an author as everything seems to point towards them being included. The key is substantive/intellectual contribution. Its extremely common for a statistician to develop the analysis plan and "outsource" the actual programming to someone junior (master's level statistician, student/post-doc). That junior person may or may not be included as an author. Just like the RA on the project who is just following orders doesn't necessarily get authorship.

Scientific writing is not like writing a novel. Authorship is not about "who put the most words on the page." The overwhelming majority of co-authors won't write a single word beyond some editing. Writing is often one of the less substantive parts of the process. If I had that level of involvement on a project and got cut from authorship, I would certainly never help that person again (and might be reporting them for ethical violations).

Thanks for your input

This is the first project I have been so heavily involved in, so it’s good to hear how it’s meant to work professionally. I would think that whoever does all of the analyses should be on the paper even if they were given a data analysis plan, but I suppose all ras don’t go on the paper for their work so I understand.

If you don’t mind me asking...who is typically the person who writes the most on a paper? For instance, in my situation, I was an undergraduate working on this project, have just recently graduated, and will be 2nd author (the PI will be 1st as always in our lab). Does all of the writing fall on me and the PI in that case? Or should I speak to the PI about what her expectations are in terms of written contribution from me?

Many thanks!!

 

[WisNeuro]

If you don’t mind me asking...who is typically the person who writes the most on a paper? For instance, in my situation, I was an undergraduate working on this project, have just recently graduated, and will be 2nd author (the PI will be 1st as always in our lab). Does all of the writing fall on me and the PI in that case? Or should I speak to the PI about what her expectations are in terms of written contribution from me?

Many thanks!!

Typically, the 1st author should be writing most of it. But, you should definitely speak to the PI about contributions to the manuscript and analyses before anything. This should be outlined in some way before anything really happens, in an ideal circumstance.

 

[Ollie123]

Who does most of the writing will vary by context. Probably most common is for the first author to do > 95% of the initial draft. Sometimes, they will write key sections and then (say) the statistician will write the results. Sometimes someone else will draft the methods section...this is often the easiest section so some folks have junior folks write this for them and then edit. All of this depends what level someone is at. A senior PI will often be first-author on the main paper from a large grant, but will write only a small part of it themselves and "oversee" folks on their team who write individual sections (or one junior person who writes most of it and is listed 2nd or 3rd). Alternatively, they may take last/senior author and let a mentee take first.

Regarding the person who does the analysis being included...it really does depend what we're talking about. If its a really basic analysis plan, it is really RA-level work to run a few ANOVAs/regressions/etc. using variables they are told to examine. Obviously, if we are talking about something that takes weeks of full-time programming for an experienced programmer to clean up the data and then figure out how to run a complex analysis, check assumptions or other things that require use of subjective judgment...that is quite another.

 

[propsych]

Typically, the 1st author should be writing most of it. But, you should definitely speak to the PI about contributions to the manuscript and analyses before anything. This should be outlined in some way before anything really happens, in an ideal circumstance.

Ok, I have a meeting set up for this week to get a better idea of my role. Thanks!

 

[propsych]

Who does most of the writing will vary by context. Probably most common is for the first author to do > 95% of the initial draft. Sometimes, they will write key sections and then (say) the statistician will write the results. Sometimes someone else will draft the methods section...this is often the easiest section so some folks have junior folks write this for them and then edit. All of this depends what level someone is at. A senior PI will often be first-author on the main paper from a large grant, but will write only a small part of it themselves and "oversee" folks on their team who write individual sections (or one junior person who writes most of it and is listed 2nd or 3rd). Alternatively, they may take last/senior author and let a mentee take first.

Regarding the person who does the analysis being included...it really does depend what we're talking about. If its a really basic analysis plan, it is really RA-level work to run a few ANOVAs/regressions/etc. using variables they are told to examine. Obviously, if we are talking about something that takes weeks of full-time programming for an experienced programmer to clean up the data and then figure out how to run a complex analysis, check assumptions or other things that require use of subjective judgment...that is quite another.

Wow, I thought it would be more of a 50/40/10 kind of arrangement more or less, not over 95%. If my PI expects me to write it 50/50 or 60/40, would that still be a reasonable request on her end? I’m happy to do my part, I just want to be sure I understand what is reasonable for me to ask for etc.

 

[Ollie123]

I think reasonable is whatever folks will agree to. You'll note I said "typical" and not "reasonable." Typical because in busy labs, it is often easier to have one person applying focused effort on a project than trying to mash together bits and pieces. Atypical is different from unreasonable.

Also keep in mind that we don't know you and aren't in your lab. You seem to be looking for an exact answer on determining proper authorship order, but there isn't one as it varies too much by field, etc. Is this a genetics paper with 250 authors? Was this a social psych paper run on an undergrad subject pool that can be cranked out in a semester with modest effort from maybe 1-2 individuals? Are you an excellent writer or is the PI really intending this as a training experience for you to do some writing with the expectation that he will have to spend more time editing your work than it would take him/her to write it themselves?

And please don't come back with answers because that is a fraction of the questions one would need to ask and not the point Really, it is going to be impossible and inappropriate for us to nitpick authorship order from afar. Like it or not, this is subjective. There is no formula for determining authorship. I would discourage you from being too obsessive about it - I know I (and I suspect most) are much happier and more productive when working as a team, focusing on getting the work done and not worrying too much about exact portioning of credit. Totally reasonable for you to talk with the PI and discuss the rationale for authorship order, as an educational experience if nothing else. Obviously, I think its important to make sure you aren't taking advantage of on these things but I'm not hearing anything that sounds at all unreasonable here, at least based on what you have presented.

 

[propsych]

I think reasonable is whatever folks will agree to. You'll note I said "typical" and not "reasonable." Typical because in busy labs, it is often easier to have one person applying focused effort on a project than trying to mash together bits and pieces. Atypical is different from unreasonable.

Also keep in mind that we don't know you and aren't in your lab. You seem to be looking for an exact answer on determining proper authorship order, but there isn't one as it varies too much by field, etc. Is this a genetics paper with 250 authors? Was this a social psych paper run on an undergrad subject pool that can be cranked out in a semester with modest effort from maybe 1-2 individuals? Are you an excellent writer or is the PI really intending this as a training experience for you to do some writing with the expectation that he will have to spend more time editing your work than it would take him/her to write it themselves?

And please don't come back with answers because that is a fraction of the questions one would need to ask and not the point Really, it is going to be impossible and inappropriate for us to nitpick authorship order from afar. Like it or not, this is subjective. There is no formula for determining authorship. I would discourage you from being too obsessive about it - I know I (and I suspect most) are much happier and more productive when working as a team, focusing on getting the work done and not worrying too much about exact portioning of credit. Totally reasonable for you to talk with the PI and discuss the rationale for authorship order, as an educational experience if nothing else. Obviously, I think its important to make sure you aren't taking advantage of on these things but I'm not hearing anything that sounds at all unreasonable here, at least based on what you have presented.

Thanks for walking me through this so much. I’m getting the idea. I will just ask my PI how she envisions it working and go from there. I’m happy getting to work on this regardless so will step back a bit and go with the flow.

 

[Justanothergrad]

I want to just mirror what Ollie has said and reinforce the idea that authorship and expected contribution varies substantially by lab. There is ethical/unethical, but outside of that it varies widely. Thus, conversations about involvement level are always important to have with PIs, colleagues, and advisees. It can be a difficult situation for a student because of power dynamics, but it is still a conversation to have. You can advocate for yourself as well during the process. So, if you are not familiar with the statistics used, you probably shouldn't be writing the results section (or, at least not without increased mentorship) and that is a reasonable thing to ask for increased assistance with in my eyes.

 

[propsych]

I want to just mirror what Ollie has said and reinforce the idea that authorship and expected contribution varies substantially by lab. There is ethical/unethical, but outside of that it varies widely. Thus, conversations about involvement level are always important to have with PIs, colleagues, and advisees. It can be a difficult situation for a student because of power dynamics, but it is still a conversation to have. You can advocate for yourself as well during the process. So, if you are not familiar with the statistics used, you probably shouldn't be writing the results section (or, at least not without increased mentorship) and that is a reasonable thing to ask for increased assistance with in my eyes.

Sounds pretty darn logical. I'll likely be sticking to the intro/discussion/and/or methods!