Collaborative datasets and their resulting publications can produce some truly spectacular insights. But as more voices find their way into the collective conversation, managing the social interplay between partners can result in confusion and lost time – the continual headaches that creep up all too often in collaborative work.
Meredith, a young research scientist, was on her way to pitch a new project idea to some up-and-coming researchers. She walked into the unfamiliar lab with a confident spring in her step; as a seasoned collaborator, she had a pretty good idea of what to expect from this first meeting and what strategies she might employ to win over the researchers and get them behind her project. At the meeting, Meredith chatted with the researchers and students she knew and introduced herself to those she did not. Once all were familiar with each other and the regular lab business drew to a close, Meredith launched into her proposal: a collaborative analysis of several datasets to look for patterns in bacterial community genetics.
Ordinarily, Meredith might expect some hesitation at the request to contribute a dataset. After all, the PhD students before her had labored tirelessly for years to collect the data in question, and they might not relish the idea of handing over their work to an acquaintance or distant colleague (however friendly). But in this case, Meredith had reason to hope for cooperation; the graduate students had all recently contributed their datasets to a publication organized by Dr. Knowles, a PI whom Meredith had worked with in the past. The experience, she reasoned, had assured them of the data sharing process and broken the ice for future collaboration. While an increasingly common practice, data sharing had been known to spook many a distrusting researcher. But as recent initiates into the world of data sharing, the students before her would have a much more accurate picture of its tremendous benefits and minor risks. And if some of them still needed a little convincing, Meredith would happily instruct them with a full account of how data sharing had transformed her own work time and time again.
Prepared though she was, Meredith proceeded with caution, carefully laying out her suggestions and then leaving the decision in their hands.
“I want everyone to feel like they can play the role that they want to play. There are going to be lots of different ways to analyze these data and chances for us to work on this together.” She briefly mentioned how everyone involved in the analyses could expect equal access and equal authorship, but opted to keep the meeting light and skim the details of how this could be achieved – after all, they would have plenty of time to discuss over email the specific plans for contributing and securely sharing the data. The details, she reasoned, could wait.
Meredith’s diplomacy paid off. As the meeting was winding down, the PhD students and their PI all expressed excitement about the collaboration, and aggregating the datasets seemed a natural next step.
Once back home, Meredith sent out a group email to everyone who had been at the meeting, thanking them for their support and reiterating the plan. But any optimism Meredith was feeling quickly deflated with the first email reply. Dr. Thackeray, the second PI whom Meredith had only met briefly, was not exactly pleased with how she had handled the correspondence.
“Put yourself in these students’ shoes. You want them to hand over the whole of their PhD work without giving them a chance to voice their opinions to you one-on-one? You can’t just ask someone to hand over their data in a group email.”
Though Meredith hated to admit it, Dr. Thackeray had a point – her strategy may have been an efficient, time-saving way of moving forward, but it risked jarring the contributors’ trust. After all, how could they expect she would uphold her word and not just use the data for her own purposes? In her work, Meredith was usually surrounded by like-minded colleagues who regularly shared data and celebrated its benefits. Meredith had perhaps too hastily interpreted the group’s enthusiasm to mean that none of the researchers at the lab would have reservations about data sharing. This unexpected response reminded Meredith that she needed to reconsider the whole affair from the perspective of a researcher who had never shared data so openly. How could the researchers know that other contributors from the same lab would not scoop the data once it became available to the group? The lab meeting had been a great way to get people interested in the idea of collaboration, but it alone was not sufficient in easing their all-too-reasonable concerns.
Somewhat embarrassed by her misreading of the situation, Meredith quickly composed individual follow-up emails to each potential contributor. She precisely outlined how the data would be used, and clarified that no one else would be granted access to the data without permission from the other contributors. Most importantly, Meredith gave everyone an opportunity to ask questions directly or voice any concerns about the process. The bottom line was clear: any contributed datasets would be in good hands. As a show of good will and a way of enticing everyone to collaborate, she even offered co-authorship to anyone willing to submit a dataset.
The peace offering appeared to have worked – datasets began trickling in from over twenty of the graduate students. But after already handing out the reward of coauthorship, Meredith had lost her one and only bargaining chip; she had nothing left to motivate the students into going beyond mere data contribution and weighing in on the manuscript. And even while many of them genuinely wanted to do their part in bringing the manuscript together, their schedules and other projects restricted how much time and effort they could realistically expect to dedicate to the cause.
So Meredith began writing and analyzing, laboring intensively and solitarily over the paper that she had envisioned working on as a group. The first draft, once completed, was entirely of Meredith’s making. She had hoped for more input from the PhD students and postdocs, and rightfully so – as coauthors, shouldn’t they be expected to help move along the analyses? Why should they receive equal authorship while she continued to perform most of the work? Contributing data alone did not warrant authorship, Meredith knew. But she had already made the promise.
If there was an easy solution to the mess before her, Meredith couldn’t see it clearly. As she continued to piece together the manuscript, her brain clicked into overdrive, playing and replaying all the possible scenarios of how the issue of authorship in the final manuscript would unfold. What if she sent around the incomplete manuscript, noting the trends she had so far pulled from the datasets, but also that there was a lot of room for further analysis, should others decide to contribute? Maybe then she could gently nudge her collaborators into taking a more active role and earning the authorship she had prematurely promised them.
Dr. Thackeray appreciated where Meredith was coming from, but nevertheless struck down her plan. “This would be a waste of time for your coauthors. They would spend all this time putting together analyses and formulating opinions, and you could not possibly include it all in one paper. And what if someone submitted an analysis that didn’t prove useful? You would have to turn them down for authorship after already making the promise.”
And so Meredith found herself forced back to the drawing board, racking her brain for another solution, another way to get the manuscript completed without overtaxing herself, but still maintaining her promise. What if she let everyone know that the manuscript was more or less complete and only required minor edits? Those who agreed to providing edits and insisted on being named as coauthor could still expect recognition as a contributor, and a second spinoff manuscript could take shape if others decided to contribute more analyses. Of course, this solution also came with some glaring imperfections. How would she ultimately decide who received coauthorship – by appeasing only those with the loudest voices, the most complaints? And how could she begin to think about another manuscript when the current one still required so much work?
Try as she might to come up with a better solution, Meredith felt stalled – and the complexity of the situation was detracting too much from the existing tasks at hand. She continued chipping away at the manuscript, aided here and there by the edits of other contributors, but the work was a mostly solitary endeavor. Though the unfairness of the situation sometimes made her start to regret ever deciding to collaborate, she had to find a way to fulfill her promise of authorship to all who contributed datasets. Instead of extending individual authorships to everyone, Meredith decided (at the urging of some of the more experienced researchers in her network) to include a group authorship. By crediting everyone under the title of Bacterial Community Genetics Lab Group, Meredith could concisely and at least somewhat fairly acknowledge the efforts of everyone involved. But should she consult the group before going ahead with her plan? Doing so would mean opening the authorship up to discussion and inviting suggestions – additional confusion Meredith wished to avoid at any cost. Instead, she received permission from the journal’s editor and went ahead with the group authorship plan. As it turned out, her worries were unfounded; everyone at the lab group approved of the group authorship decision. The issue that had kept Meredith awake at night for weeks on end was finally resolved.
The experience was far from what Meredith had envisioned, but her frustrations did produce something valuable besides the manuscript: a resolve to improve the process in future collaborations. After working on so many collaborative projects in the past, Meredith could never have anticipated so much confusion over contribution and authorship. But as a veteran who had earned her first battle scar, she would begin future manuscripts with clearly established policies and agreed upon expectations; only active contributors would gain authorship, and submitting a dataset, no matter how complete or essential to the whole, would not be enough. Perhaps perfect solutions to the collaboration conundrum did not exist, but she would press the limits until she hit upon something that worked for everyone – only those that contributed to the process and did what they could to see the manuscript through would earn the reward of authorship.
Image: CC BY-NC-ND 2.0 by Ernst Gräfenberg via flickr