Friday, November 7, 2014

Comment: Providing better training and better career perspectives to graduate students

The fact that the way research is currently conducted in universities is not made optimally in terms of cost-efficiency and of student formation / career perspectives is something which is rather obvious, and which has been bothering me since some times. Thus, it was pretty interesting for me to read this opinion paper published in PNAS.

Alberts B, Kirschner MW, Tilgham S, Varmus H (2014) Rescuing US biomedical from its systemic flaws. Proc Natl Acad Sci USA, 11:5773-5777.

It is part of a larger debate (and it even is the target of some discussion even in LinkedIn groups (for instance the AAAS group). I would like to comment on two specific points of the paper.

First, the need to downsize academic research labs. I believe since already some times that the current model which promotes the existence of very big labs is not the best. First, while the senior scientist can direct numerous projects, especially in the same area of expertise, our time is still limited. Meaning that we spend more time doing administrative things than actual science. This also means that the time devoted to each trainee will be way smaller. When you have simultaneously 10 students and 5 postdocs in your lab, it is not reasonable to believe that you will give 1h of individual attention per day to each of them. Well, sadly, even 1h of individual meeting would be probably difficult, since we have quite a lot of other tasks to do as well. Universities want to have more and more student graduating, so we have in terms of career some interest to manage large labs. But what about the quality of the formation we provide? And of the mentoring we do? Not areas of research are the same. In molecular biology it is very common to see “hives” of students working in the labs, since a lot of the experiments follow very strict protocols, and maybe (and the maybe is important, as I am not that sure of that actually) requires less guidance by the supervisor. However, in the field I know the best, which is behavior (or cyberbehavior, but it is just a specific case of behavior in general), while the protocols are usually simple, what makes the difference is how you exactly do it. How you design the protocol, how you quantify the actual behavior, how you control for any bias not only when developing the experiment but also during the process (since behavior is a dynamic display). And more importantly, we need time and intense contacts to really “transmit” not only knowledge, but also our own experience to our students, so they can really learn and benefit from it. That is particularly true for a PhD student or a post-doctoral fellow. This type of relation is a one/one relation, where the word “mentor” takes all its meaning. However, there is a second element favoring the fact that big labs are growing. And that is an economic argument. A student costs considerably less than a regular employee. So it is easier to have a lot of students in the lab rather than a lot of workers. But are students only cheap workers? I would think no: first they are still learning, so their productivity is not at the best (and it shows, since the larger labs obviously have more publications than smaller labs, but often the number of publications per capita is lower). And second, and more important, they are learning skills to get a job later on. And here comes the issue that we are maybe forming too many students in some disciplines that what the needs really are. A lot of public funding is thrown to research for Alzheimer’s disease, which is rightly a major problem of public health in Western countries. However, these funds are used to form a lot of students in this topic. Most of them would love to do an academic career. Would it be reasonable to believe that universities will recruit every year dozen of newly formed researchers just on this topic? Of course, not. On the other hand, we are struggling to find specialist of research in the field of rhinology, while we do have teaching need for that, and while disease related to the respiratory system and superior aerial ways are critical as well. Downsizing the labs to more reasonable and manageable research units really centered on the professor research interests, and allowing a better, deeper form of trainingship for graduate students.

Which leads to the second point. We still need people to do the research: nowadays research is getting more and more complex and complicated, interdisciplinarity is a key word. Modern research relies on various expertises. Staff Scientists could be a solution, as suggested in the paper. It would increase the quality of the research done. It would provide some new positions for all the students we are forming. It would increase also the formation of the students by having more senior people around them. But there are still a few issues with that. First, such a solution would have a cost. We can not expect recruiting high qualified people without decent salaries. This would have to come from the funding we get ... while the actual tendency for research grants is more oriented toward decrease, and the current success rates are falling down. Second, the nature of the employer might be a problem. Would it be the researcher? Meaning that, if a researcher loss his grants, then the staff scientist would be fired? It would be very difficult to secure on the long term this key people. Would this people be hired directly by universities? While that might be an acceptable solution for research institutes, it is hardly something which can be defended in the context of universities. Also, not being the actual employer would mean that it would be the university who will decide with whom (with which researcher) the staff scientist will be working. With all the problems which can arise for that. And with the issue that they would be considered mainly as super-technicians and would see their intellectual contribution potentially decrease. A possibility could be to develop new types of grants, which would be accessible only for staff scientist, under the direction and projects of university researchers (but independent from the professor own grants), with a system of renewal akin to the system we have for career grants in biomedical research in Quebec province. It would of course some money, but the results for the efficiency of both academic-based research, and more importantly, student formation, might be worth the deal.

Friday, October 17, 2014

Comment: Star Wars, X-Files, Zombies, and Mermaids

Star Wars? Zombies? Steampunk???? Common, you are gotta kidding, it can not be real research!

Well, while I please myself to think I am quite humorous (I know that it is the matter of heated debates in my laboratory), I am in fact not kidding. For those interested in studying virtual communities, virtual communities related to science fiction or fantasy universes are of the highest interest. Why so? There are quite a few reasons to that. I will not go over them here; if you are interested by that, I will gladly refer you to our papers.

Guitton MJ (2012) Living in the Hutt Space: Immersive Process in the Star Wars Role-Play community of Second Life. Computers in Human Behavior, 28:1681-1691.
Guitton MJ (2011) Immersive role of non-required social actions in virtual settings: the example of trade role-play in the Second Life Gorean community. Design Principles and Practices: an International Journal, 5:209-220.

But in fact, if we go back to the “good ol’ time” (and in the field of cyberpsychology, a few decades are already an eternity), it all began there. Let’s take one of the papers considered as seminal in the field of study of virtual communities, the famous paper of Kozinets 1997.

Kozinets RV (1997) I Want to Believe: a Netnography of the X-Philes’ Subculture of Consumption. In Brucks M, MacInnis DJ (eds.): NA - Advances in Consumer Research Volume 24. Provo, UT : Association for Consumer Research, pp. 470-475.

Back then, Kozinets was interested in the structure of a virtual community in a eMarketing logic (or what would later on be referred to as eMarketing). For that, he by the way had to define a new methodological approach, which he named “netnography” (InterNET ethnography – a term which got quite a successful fate, although now the term virtual anthropology might be favored). And which community was he studied? The community of the fans of X-Files! Yes, Agent Mulder and Agent Scully…


Scully, do you believe?

No offence intended to X-Files fans… but it is not what we immediately associate with rocket science. Nonetheless, the work of Kozinets was instrumental in developing the field of eMarketing (and is rightly acknowledged as so).

Similarly (at our modest scale), researching the virtual communities of Star Wars and Gorean role-players in Second Life helped us to unveil some really fascinating mechanisms underlying the structuring of virtual communities at large. And studying the communities of players of Zombie apocalypse games provided very interesting insights in the way moral dilemmas are resolved in extreme survival situations.

Cristofari C, Guitton MJ (2014) Surviving at any cost: Guilt expression following extreme ethical conflicts in a virtual setting. PLoS ONE, 9:e101711

While the X-Files fan community, the Star Wars Role-Play community, or the Gorean communities refer to imaginary settings which follow a defined set of rules (a “canon”), some science fiction or fantasy communities are not united by a predetermined and restrictive corpus. That’s for instance the case of the Steampunk community, or of the Mermaids community in Second Life. However, these virtual communities are still potentially providing us a lot of information. For instance, studying the Second Life Steampunk community allowed us to develop a new tool to “map” the perceived distance between communities by quantifying their aesthetics (paper just accepted a few days ago, soon more on this blog!).

So, yes, even if it might look surprising, this is real research. With the fun as a bonus.


A mermaid in the seas of Second Life.

Thursday, July 24, 2014

Journal Metrics - Computers in Human Behavior

Well, we just got the new Journal Metrics today, and once again Computers in Human Behavior is doing really well ! In fact, the impact grew significantly for all of the markers.

If we take the SNIP (Source Normalized Impact per Paper: an impact factor weighted by the relative size of the field):
SNIP 2011: 2.133
SNIP 2012: 2.151
SNIP 2013: 2.408

The SJR (SCImago Journal Rank: a global prestige evaluation) follows the same trend:
SJR 2011: 1.577
SJR 2012: 1.589
SJR 2013: 1.791

The IPP (Impact Per Publication: which shows the ratio of ratio of citation per article published) is even better !
IPP 2011: 2.826
IPP 2012: 2.908
IPP 2013: 3.281

Which is likely to be reflected in the up-coming more conventional Impact Factor.
Sure, all of these metrics are just proxy (such as the Impact Factor, by the way), but still, they show us that the journal is doing really well, and that we are confirming its position as a recognized leader in the field.

Saturday, July 12, 2014

Lab Update

The last months have been rather busy, which explains somehow why I haven't been posting since quite some time. Well, time now to catch up with what happened recently in the lab! Today's post will deal with the meetings in which the lab was present in the last few months, and I will talk about our recent publications in another (latter) post.

In the last few months, we went to three really fascinating meetings, the PCA/ACA Meeting, the RAI Meeting, and the FENS Forum.

The 44th PCA/ACA National Conference (Popular Culture Association/American Culture Association) was in Chicago, IL, in April. Like always, it was an extremely interesting meeting for me. I love the PCA/ACA meetings for the incredible fresh ideas I get there, stepping from the so diverse approaches and points of view that are displayed there. Cécile Cristofari was also presenting some of our work on the Steampunk community of Second Life there, and got quite a nice feedback.

The Royal Anthropological Institute 2014 Conference: Anthropology and Photography, in the British Museum in May was as well an incredible event. I was the Convenor and Chair of the first session, which was entitled "Anthropology and photography in the digital age". Both Cécile Cristofari and myself gave a talk in this very symposium. We had an amazing feedback, and I was extremely surprised (and pleased) to see how much modern anthropology has incorporated cyberspaces in its practice. As well, I was (again) both surprised and pleased to witness the fact that Computers in Human Behavior had quite a good reputation among those anthropologists interested in the study of social media and other virtual worlds. Beside being in such a prestigious location than the British Museum, the quaility of the talks were mostly of the first order, and we came back with a lot of new contacts and ideas. I even wrote an editorial in Computers in Human Behavior about that (going to be out soon, hopefully). The meeting was preceeded by a few days that I spent at the Future of Humanity Institute of PRof. Nick Bostrom at the University of Oxford. A very interesting experience with a lot of inter-discplinary interactions, very valuable for me, and which elicited quite a few new ideas and angles of view for me.

Introducing the "Anthropology and photography in the digital age" Session
at the RAI Conference at the British Museum

Finally, the 9th FENS Forum of Neuroscience (Federation of European Neuroscience Societies) just took place last week in Milan, Italy. The FENS Meetings are always among my favorite ones, due to a few things, including (but not limited to) the fact that I get to see a lot of my old friends, the fact that peple are somehow less worried to show really innovative and not-yet-published studies (in contrast to the SfN meeting where people usually only show things ready to be published, or actually already published), and to the diversity of approaches in neuroscience that you can see across the different European countries. Now that what I do is slighlty different from purely molecular neuroscience (Well, I always was mostly involved in behavioral neuroscience anyway, but now the focus on interest has clearly drifted to cyberbehavior), I was expecting that what I was doing now would elicit considerably less interest that before among my fellow neuroscientists. Surprisingly for me, that was absolutely not the case. Maybe I got slightly less people at my poster (which was on moral decision making using zombie apocalypse massively online role-playing games as a model, the related paper just got published in PLoS ONE a few days ago) than when I was doing animal model of tinnitus, the quality of the feedback was amazing. I got people who were really interested in some of the various aspects of what we did: virtual spaces, cyberbehavior, decision making process... Who knows, maybe some interesting collaborations might follow?

The entrance of the FENS Forum, at the center of congress of Milan.

Our poster, conveniently located close to an emergency exit...
with zombie invasion, you never know.

Which confirmed my initial observations that scientists are in fact big children.

Computers in Human Behavior flyers at the Elsevier booth.