User:Daniel Mietchen/Talks/COASP 2010/Notes

Background

(CC) Image: Public Library of Science
Do journals provide sufficient contextualization for research?

• For technical reasons, publishing was historically a separate step, performed about once per iteration of the research cycle

• Publishing every relevant bit of information immediately at each step is technically feasible now, and the remaining hurdles are cultural ones.

• Wikis allow for systematic linking and thus enhanced contextualization (sidenote: some have argued that links are distracting)

• Overview of the evolution of wikis and wiki-like environments

Mentions MediaWiki plugin for Wordpress

Etherpad

Google Docs

• Dear Scientists, Stop Duplicating Work

Wikis as platforms for science communication

• Wikis can be used, in principle, for any aspect of scholarly communication, as detailed in this comparison of wiki- and paper-based communication systems and the related blog post.

• Examples exist for all steps of the research cycle, except successful applications to major funders (see this overview for some attempts)

• Benchmark: English Wikipedia

Top 10 Reasons Why Academics Should Edit Wikipedia

Wikis as platforms for scholarly publishing

• The idea is not new — WikiSciencePublication stated in 2006:

"Somewhere at the fringe of science, someone will start using wiki publishing for science publishing."

See also Wikiversity:Publishing original research

• Most variants of wiki software can handle user rights in great detail, so as to allow implementation of basically any kind of peer-review system

• Lab reports: Lab reports

• Conferences: Stand-alone site / contextualized schedule / talks, contextualized talks and Posters (possibly also as clickable imagemaps, like here), presentations

• Research papers: Accompanied by wiki article (progress report), contextualized on-wiki, links to wikis amongst other sources, integrated into database

• Review articles: Scholarly / Outreach

• Wikis can act as publication filters for new research

See also Can Computers Help Scientists With Their Reading?

• Books

• References

• Publication lists (incl. supplementary materials and in principle direct links to the raw data)

• To-do lists for a whole discipline

See also CoLabScience

• built-in article-level metrics at bottom of page and via What links here (which could also be used for other pages, e.g. those hosting images, references, or datasets), author-level metrics via Special:Contributions, further aggregation possible (e.g. at the level of research projects, labs or thematic workgroups)

• Wikipedia Journal

• Knol shares some aspects with wikis and blogs and is already in use for PLoS Currents.

Wikis as platforms for Open Access publishing

• The majority of wiki platforms are open access by default

• Stresses the re-use part of CC licenses (e.g. for images) — an aspect of OA publishing that does not receive much attention outside research blogs (cf. detailed discussion with respect to the American Physical Society, arxiv and Quantiki, and the final outcome: APS authors keep copyright over derivative works). While the main purpose of such licensing is certainly to make the research available to other specialists working in the field, they also make the research available to scientists from other fields (not just in academia) as well as non-scientists (teachers, entrepreneurs, media, patient groups or hobbyists).

Examples: Gyrification, Surface-based morphometry and Chordoma

Also note that Fig. 3 of http://dx.doi.org/10.3389/neuro.11.025.2009 and Fig. 2(III) of http://dx.doi.org/10.3389/fnins.2010.00020 explain the same thing, original to neither papers

• Incompatibility of wikis and closed-access publishing

re-use of CC-BY-SA content in copyrighted papers

Quality assessment

In principle, any system of peer review can be implemented on a wiki, by detailed management of user rights: The usual single-blind as well as double-blind or open peer review, with the reviewers or even authors always or optionally, temporarily or permanently remaining anonymous, with simple accept/ revise/ reject decisions or interactive two-stage or multi-stage discussions, in public or hidden from it (possibly even in part), before and/ or after formal publication.

Some wiki examples:

1. English Wikipedia (Flagged Revisions, WikiTrust, Wikipedia:Pending changes)
2. Scholarpedia
3. Encyclopedia of Earth
4. Citizendium

Some non-wiki examples:

1. Copernicus journals
2. PLoS journals
3. Frontiers journals
4. BMC journals
5. Semantic Web journal
6. Rejecta Mathematics
7. WebMedCentral

Business models

• Main ones: author-pays, (partial) subscription, philanthropy, advertising, premium services

Opportunities

• Article-specific Job ads (e.g. via subpages)
• Wiki for Calls For Papers
• Post-publication peer review: Faculty of 1000, The Third Reviewer, Rejecta Mathematica
• Wiki export, or standardized XML or HTML output that could be imported to a wiki via some xml2wiki or similar converters
• Image search (example) and annotation
• Search by license (prototypes: journals, images)

PD Image
Search by license — not possible yet. Why?

• Semantic integration

Non-wiki example

MediaWiki as a blog, using Semantic MediaWiki

• Integration of non-text media with text (just like images; non-wiki audio example)

Also for references

• Automated pre-population from databases

Notes

• Essential elements of science publishing:
  • Research
  • Documentation
  • Making things public
  • Integration with previous and future knowledge
  • Discussion

• Seven quality indicators to strive for in an ideal peer review system (per Jefferson et al., 2002)

1. importance,
2. usefulness,
3. relevance,
4. methodology,
5. ethics,
6. completeness
7. accuracy of an article
8. accountability (added by Peter Frishauf in part 4)

---

• use screenshots of Recent changes from mainspace around midnight from WP, EoE, SP, CZ, WE & OWW & Knol

---

• Problems from I hate your paper:

1. Reviewers are biased by personal motives
2. Peer review is too slow, affecting public health, grants, and credit for ideas
3. Too many papers to review

• Respective solutions:

1. Eliminate anonymous peer review ( Biology Direct, BMJ, BMC); run open peer review alongside traditional review (Atmospheric Chemistry and Physics); judge a paper based only on scientific soundness, not impact or scope (PLoS ONE)
2. Shorten publication time to a few days (PLoS Currents Influenza); bypass subsequent reviews (Journal of Biology); publish first drafts (European Geosciences Union journals)
3. Recycle reviews from journals that have rejected the manuscript (Neuroscience Peer Review Consortium); wait for volunteers (Chemical Physics Letters); reward reviewer efforts (Biology Direct, BMC, Frontiers, ACP)

---

• Mention possibility of linking to Open Access funding, as per Microfinancing for science

---

• Liquid Publications — the editable web

---

On the cover image: Journals are not essential to a wiki system described here, as their function of conveying new information would go away, but they still have their role in turning edit histories into stories, and in framing debates, for which wiki talk pages are not ideal.

---

Mention alternatives to wikis:

• Etherpad,
• Google docs
• ScribTeX
• Buddypress

Quotes

From the JoPM peer review discussion (my comments)

Peter Frishauf, founder of Medscape (3:29-3:44 in part 4):"Wikipedia is probably the most robust Petri dish we have for actually studying the process of words and contributions, because it is auditable."

Elizabeth Wager of Jefferson et al., 2002 (10:37-10:56 in part 4): "So, we had the idea that you do your systematic review before you do your research; you do your research, and then if you haven't changed much, you haven't really made a big impact, whereas if you've actually shifted things one way or the other and made it more precise then you have."

Peter Sefton (The next wave in scholarly word processors?): "what I think we need in scholarship is the web, but editable"

Chris Gutteridge (in 1st comment on Sefton's post): "Better still, if you assert something said in another paper, sod the citation, transclude the relevant text, with a full electronic citation allowing you to verify it."

From I Hate Your Paper - The Scientist "I believe strongly [that] in the end, all life is on the record," Smith says—"you should stand by what you say, and you should put your own name on it. It makes me uncomfortable that science has moved away from that."

• SPECULATIONS ON THE FUTURE OF SCIENCE By Kevin Kelly : "Wikiscience" is leading to perpetually refined papers with a thousand authors. Distributed instrumentation and experiment, thanks to miniscule transaction cost, will yield smart-mob, hive-mind science operating "fast, cheap, & out of control." Negative results will have positive value (there is already a "Journal of Negative Results in Biomedicine"). Triple-blind experiments will emerge through massive non-invasive statistical data collection--- no one, not the subjects or the experimenters, will realize an experiment was going on until later. (In the Q&A, one questioner predicted the coming of the zero-author paper, generated wholly by computers.)"

From Speculations on the future of science:

• "Wiki-Science – The average number of authors per paper continues to rise. With massive collaborations, the numbers will boom. Experiments involving thousands of investigators collaborating on a "paper" will commonplace. The paper is ongoing, and never finished. It becomes a trail of edits and experiments posted in real time — an ever evolving "document." Contributions are not assigned. Tools for tracking credit and contributions will be vital. Responsibilities for errors will be hard to pin down. Wiki-science will often be the first word on a new area. Some researchers will specialize in refining ideas first proposed by wiki-science. "

• "New ways of knowing will emerge. "Wikiscience" is leading to perpetually refined papers with a thousand authors. Distributed instrumentation and experiment, thanks to miniscule transaction cost, will yield smart-mob, hive-mind science operating "fast, cheap, & out of control." Negative results will have positive value (there is already a "Journal of Negative Results in Biomedicine"). Triple-blind experiments will emerge through massive non-invasive statistical data collection--- no one, not the subjects or the experimenters, will realize an experiment was going on until later. (In the Q&A, one questioner predicted the coming of the zero-author paper, generated wholly by computers.)"