Publishing Socio-tech work
From Sociotechwiki
Contents |
[edit] Overview
Publishing Socio-technical Work --> Weisband
Packaging work
- Picking an audience
- Co-authorship
Strategy of publishing
- Stream of research
- Portfolio of publications
Issues
- Different things count in different ways: peer-review publications, books, conferences
Value of discussion
- Concrete discussion are helpful; was good to find out what was going on in other institutions.
- Develop relationship: need to do more than just conference publications; need to other things, too, to meet people and network within the academic community (to develop tenure letter writer relationships).
- Have a research stream: Not as much about the number of publications as about the theme, how you can talk about them. Show that you are building in your work to something.
- Be known for something: In many fields easy to publish a lot (e.g., due to proliferation of conferences and publications), but also important to "swing for the fences" to have something really impactful (and at top journals).
How to balance speaking to multiple community?
- Know how to speak multiple languages.
- Know how to translate.
- Be strategic in publishing in multiple areas.
- Pick carefully because of increased cost due to presence costs.
- Hour glass: job talk time (be flexible); pre-tenure focus on something to be known for; later broaden out again.
- Aim for cross-pollination and building bridges (instead of split personalities)
How to you get other people to give you feedback?
- Okay to send request for feedback to someone who might be a future reviewer.
- Good strategy is to ask for specific feedback.
- For the really busy people (like journal AEs and NSF program officers) is to send a one page summary.
- For doctoral students, you are probably more likely to get a response if the request comes from the adviser rather than student (at least the first ever communication).
- Make it easy for the person you are asking for help: send one page, ask questions in one part, provide multiple ways to respond (a phone number, offer to meet at a conference, in person if you're going to be nearby). Many people really want to help, make it easy for them to.
- Usually, a non-response means someone is just really busy. One exception: at NSF, it is a requirement to respond to emails (it is a performance guideline). It is appropriate to follow up with them if no response to ask "did you miss this?"
[edit] Management and Organization
[edit] CHI/CSCW
CHI/CSCW Community Report on ST Research
The approach that ST Research takes in the CHI/CSCW communities is somewhat different from the other communities in that it includes a focus on design and the design process. We brainstormed about papers/concepts in four large categories: Theoretical influences (aka world views or abstractions), methods, domains examined (though not limited to them), and technical designs.
A. Theoretical influences
a. Distributed Cognition (Hutchins)
b. Activity Theory (Nardi, among many others)
c. Ethnomethodology (?)
d. Social psychology, especially theories of motivation, conversation
e. Legitimate peripheral participation
B. Methods
a. Ethnography – “Learning from Notes,” Orlikowski
b. Technomethodology – Dourish
c. Surveys – Kraut & Kiesler, Homenet
d. Experiments – Olson & Olson, Shape Factory, trust
C. Domains
a. Opensource
b. Wikipedia
c. Control rooms
d. Air traffic control
e. Twitter use in teens
D. Technical Designs
a. Suggest Bot
b. UARC/SPARC collaboratory
c. Babble
d. Movielens
e. Password security
f. Group undo in shared editors
E. Overviews with a point
a. Grudin – “Why CSCW applications fail,” “…succeed.”
b. Olson & Olson – “Distance Maters”
Rankings of Publishing Venues
Journals
A TOCHI A HCI A Journal of Applied Psychology A JCSCW B IJHCS (used to be man machine systems) B Information Technology and People B Interacting with Computers B Personal and Ubiquitous Computing B/C Behavior and Information Technology C IJHCI
Conferences
A CHI A CSCW A DIS A Ubicomp A WWW B Group B eCSCW B ICWSM B WikiSIM B Pervasive B Interact C IEEE SocCom C RecSIS C Comm and Tech C CHIMIT C Coop C CTS (but A for Army) D HCII D iConf (but rocketing up) A-F HICSS (depending on track)
Online
First Monday JCMC
Magazines (broad readership)
Interactions CACM
Order of authorship
Portion of contribution, first most Student first Lab manager last Alphabetical (not so common) Who wrote the first draft is first author Some times contributions are annotated.
Citations
Google scholar ACM DL H-Factor ISI (but doesn’t have computer science!)
[edit] The "nook" group
I didn't write down the names of attendees, alas.
On general venues:
We did, however, conclude that the field looks at journals and conferences about equally, depending on where you're at in terms of both institution and department. In general, departments where HCI/CSCW are a big part of the game tend to appreciate conferences more, although there is sometimes pushback at higher levels in the academic hierarchy that don't understand why there are no journal publications. Mark Ackerman's advice at last CSST: publish a few journal articles, so at least the higher levels know that you can play the game. Christine Halverson said that it doesn't matter so much in industry between journals and conferences. There was general agreement that books and book chapters are fairly low value, especially for early-stage faculty pre-tenure, because it's hard to show impact especially given the time it takes to publish a book. No matter what
On specific venues:
We didn't do a lot of ranking. There are some clear "first-level" conferences, mostly because they are the big ones that everyone goes to: CHI and UIST for the HCI folks, CSCW and ECSCW for the CSCW folks. Then there was a moderate number of second tier places; in HCI these tended to be specialized to particular sub-areas: IUI for intelligent user interfaces, DIS/DUX for design, Ubicomp for mobiles. In CSCW there seemed to be fewer specializations and "lower" conferences just tended to be smaller: GROUP and CoOp. Some conferences, such as IWCSM and Creativity and Cognition, seemed to be on the rise. Large but unselective conferences such as HCII and many tracks in HICSS were regarded as less valuable; regional conferences of the OzCHI, NordiCHI, GI (Canada), British HCI were also just publications.
For journals, we mostly mentioned the major ones: ToCHI, HCI, IJHCS, JCSCW, and (maybe) JCMC. Some publications have good names but a more magazine-like feel: CACM, Interactions, IEEE Computer. There was little love for online-only venues such as First Monday. Again, these may be place-dependent; many people at Cornell Comm like HICSS, participating in reasonable-quality tracks. Some places and disciplines probably like First Monday a lot better.
On effective strategies:
Number one was to get feedback, early enough in the process that you can do something about it. Don't wait until the last few days to get an outside read, because there's not a lot of value that can be generated from that. Getting feedback early might be especially valuable for doing cross-disciplinary work where you're looking to publish in a field that's new to you. The intellectual benefits should be clear after the conference, but from a more tactical, publishing point of view, articles that don't cite much of a particular conference, journal, or field's literature tend to look funny in terms of what they focus on and how to talk about things.
Appropriate targeting was also called out. Knowing what kinds of things ToCHI vs. HCI, or CSCW vs. ECSCW vs. GROUP, values, can help you make appropriate decisions about venues that are most likely to appreciate and be impacted by the work. Talking to people experienced in those venues can help you make good decisions (as Gilly found out when talking to Judy about the suitability of ToCHI vs. HCI for an article she's writing). Further, targeting the message _at_the_audience_ was also called out as important. Not just speaking their language, but helping them understand why what you're working on is important and valuable, and justifying and spelling out your arguments in ways that are accessible. For example, a paper about a novel system design for problem X will want to emphasize different things if the goal is to publish it in CHI (theoretical rationales and connections to other literature; often an empirical evaluation; an important or interesting domain) than UIST (descriptions of technical detail and design tradeoffs; ability to replicate; novelty and potential value). Those characterizations are stereotypes of those two conferences' review desiderata, and not everyone will agree, but the stereotypes don't just come from thin air. Personal note: perhaps a really good system paper should strive to do all of these things!
There's also the curious matter of "notes" in HCI, which are 4-page papers, compared to the normal 10-page full paper format. In some cases, notes are reviewed as small but complete contributions, "little baby papers" that should do everything a full paper does, just not in as much detail or about a topic that's more incremental. (CHI seems to have this model, often.) In others, notes are contributions that do parts of the process well, but not all of them: a system with a thorough justification but no evaluation yet might be a reasonable thing to publish; a strong description of a case study/observation of practice, but without as much connection to literature or arguments toward new designs and/or theories. (People mentioned that CSCW notes are treated a little more like this.) It's not clear that reviewers really agree on how notes should be evaluated, and this makes them confusing.
We started running out of time at this point. We did talk a little about authorship, and the "bimodal model" where lead authors go at the front, advisors at the back, and people in the middle did some work but were clearly not the main driver, worker, and/or writer -- and usually, the first author is all of those things, at least to some extent. Co-equal work is handled badly; some venues use alphabetical ordering; some advisors choose to be listed first. General advice: figure this stuff out up front, as early as you can, so everyone knows what the roles and plans are likely to be. Last minute changes, especially deletions, often lead to hurt feelings, but some of us have seen bitter disputes about the order of the third and fourth authors on a paper. As far as who's included as an author, we really didn't discuss. DanCo proposed "if you did significant work, and can explain the paper", based on some uncomfortable experiences interviewing potential grad students who couldn't talk well about papers they were authors on; whether someone who did a lot of implementation, data collection, coding of qualitative data, etc. is an author is sometimes contentious (Gilly asked about this at the very end of the session but we didn't get to push on it), and will depend again on field, department, and personalities involved. Again, general advice: figure this stuff out up front.
[edit] STS/SSS
[edit] Media/Comm
General: In media/comm, publication expectations vary wildly on where you are. Know your context.
Books: Unlike several other contributing fields to socio-tech, books can matter a lot. If they do, some of the best presses include Chicago, Oxford, NYU, MIT, Yale, Harvard, etc.
Journals: expect multiple possible orderings for the following. Here is one typical ranking as we imagine it.
- 1. Journal of Communication (Max Parks, new editor, friendly to this kind of work)
- 2. Communication Research
- 3. Media, Culture, & Society
- 4. Critical media studies and culture
- 5. Journal of Computer Mediated Communication
- 6. New Media & Society
Others outside the field:
- (STS) Science, Technology, and Human Values
- (STS) Social Study of Science
- Information, Communication, Society
- Information Society
- Journal of Social Personal Relations
- Journal of Person. and Social relations
- Journal of Information Technology and Politics
Author order: To determine author order, have an explicit conversation beforehand, and in the lack of other solutions, consider favoring whoever carries the work.
Tips for promotion/advancement: consider drawing on these sources to help make your tenure case: impact factor, h-factor, ISI, Google Scholar, Web of Science/ERIC citation alerts, syllabi tracing, most downloaded. Advancement often also depends on social relationships. Know what is expected of you and get along.
[edit] Social Informatics
Overall Points
- Right now, there are three main journals which can be though of as overlapping but distinct bubbles, each speaking to an intellectual community. To choose between them, look at who you like and where they publish.
- One way to get to know a journal is to review for them and ask for copies of the blind reviews. Write the editor or an associate editor.
- Read the journal for the past five years.
- All of these are from outside the main journals in traditional disciplines such as communications.
- You can write the editors with paper ideas to get their take.
New Media and Society
- Some communications, some sociology, some information science, but lots of new media people with communication.
- Focus on ethical and critical theory.
- A lot of discourse work.
- A lot of online processes.
- Quickest review process.
- Lots of AOIR people look at that journal.
- More European focus.
Information, Communication and Society (ICS)
- Much more a sociology journal.
- CITASA, part of the American Sociological Association, gets a special issue in that journal every year.
- Uses macro-or-micro social techniques.
- Interested in critical analyses, testing assumptions.
- Have to know that literature. Hard to cross into that one without due diligence.
- Probably 20-25% acceptance.
- Steve can't detect a pattern.
Information & Society
- Publishes mostly social analyses of computing.
- Issues around social change and technology.
- Strong communication focus.
- Look at large scale, meso, trans-organizational.
- Current editor is telecom and policy angle.
- Pressure for measures people to show up.
- 7-8000 words, hard for sociologists.
- Publish from Information Sciences.
- Published some STS-like things, maybe that didn't land well in HTSV.
- Acceptance rate around 20%.
- Most positivist.
Others
- First Monday a bridging journal, good for ideas but not theory.
- Books matter for more sociological and ethnographic work.
Conferences
- Vehicles to socialize.
- Totally ambiguous whether you credit for it.
- In journal and book based communities, small workshops are the places to go.
- Go to the professional development workshop ahead of time.
back to 2010 Schedule
