HCI Venues: A Question of Authorship
What does collaborative authorship look like in human-computer interaction venues
Human-computer interaction (HCI) is a rather broad, interdisciplinary field. I, myself, am in a unique postdoc where I get to work with and co-train behavioral scientists while doing research into how we can design technology to support people with mental health concerns.
In the process of this co-training, we have had in-depth discussions as a group about what different publication venues look for in a paper and the norms of each community (and sub-community). A question of authorship naturally arose from these conversations: who “qualifies” as an author on a paper? what are the strategies for assigning authorship? what is the best number of coauthors on a paper? does any of this change if you are early career and going on the job market in the fall?
Being an HCI researcher, I decided to take to social media and my various networks to see what other people thought about these questions. I was surprised to not find many resources on the topic readily available online, so here I am writing a blog summarizing my findings. (If you’re into numbers, Jofish Kaye has a great statistical analysis of CHI through 2008 here and rumors are he’s working on an updated version.)
I asked, “Hi my HCI friends, I am having a friendly intellectual debate with a lab mate about number of authors on conference papers (CHI, CSCW, etc.) specifically for *early career* HCI folks. Any good tips / advice / recommendations from any of you??”
Across three social media platforms, I had 20 people respond from other early career researchers to post-tenure. While this is an interesting set of data, please remember that your mileage may vary. Different institutions have different rules and norms. Requirements for types of publications and authorship will also vary by location.
So — what did I find?
Being first (and sometimes second) author counts
Overwhelmingly, respondents said that having some first author publications was important when looking at job applicants. Having first authorship shows that you are able to show some deep engagement in a research area and that you can lead that research. In fact, some of the people who responded said that the only time a detailed review of authorship and publications came up in search committees was when an applicant did not have any first author publications on their CV.
But! Being an Nth author is important too…
While it is important to have some first author publications, many people also said that having some papers with others shows that you can collaborate. Do you work well with others? Having a number of coauthored publications where you are in a variety of roles can show this. Where you are in the author list, however, after slots 1 and 2, really doesn’t matter.
Some people said that having only first authored publications could be a red flag to a search committee. Again, this isn’t just a sign of someone who can or cannot lead a research project. HCI is a very collaborative field and if you don’t play well with others, that isn’t going to be great in the long run. Again, have a variety of publication types and positions within the research project overtime, shows that you are a diverse researcher ready to take on new projects and can work in a range of research environments.
So, how many authors should a publication have?
The answer to this was, for the most part, as many as you need. Being generous with authorship helps newer researchers learn how to write better papers and conduct their own research in the future. We are, as I said earlier, a very interdisciplinary field, so you’ll also find a number of coauthors who may specialize in something other then HCI authoring publications in HCI spaces. People who responded to my queries also included those in high school and undergraduates as authors when they were also involved in the projects.
One respondent linked to the ACM Guidelines for authorship, which state that the author must have made a “substantial intellectual contribution” to the work. I think as more authors are added to a paper, this area becomes quite murky. However, of the 20 people who responded to my question, none seemed concerned about this.
Be aware of norms
HCI (and subsequently the sub-communities with HCI) had its own norms just like any other research community or field. People mentioned things like advisors being positioned last in the list of authors. Others mentioned trying to align the list of authors by amount of contribution. These things mostly hold true across HCI, but do vary from lab to lab. It’s always fun to check out how other labs handle things like authorship — gives you a sense of how they might conduct their research and train their students. Norms, of course, could be its own blog entirely, so I’ll leave the rest alone for now. Feel free to comment if you have any thoughts!
In sum, authorship is one of those tricky things that has to get worked out from project to project, lab to lab, and changes as your career develops. However, here are the key take-aways:
- Be kind, be generous, be inclusive.
- Show your diversity as a researcher by taking up different roles across different research projects. Have a strong first-author project, but also be a good collaborator and take a supportive role in other projects.
- The number of authors on a paper mostly does not matter. Adding more authors does not diminish the paper’s value or the value of being an author on that paper.
What are your experiences as an HCI researcher? Feel free to leave a comment here or reach out through social media!
Acknowledgments: Thank you to everyone in my social networks who took time to answer my questions about authorship!
