You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
[New Blog Post] SKY: Towards Science 2.0
I'm launching SKY, a tool leveraging the digital revolution to make science more Accessible, Open, Collaborative and Effective.
https://chabefer.github.io/SKY/
SKY aims to be @Wikipedia of science.
https://economistjourney.blogspot.com/2019/03/sky-tool-for-decentralized-accumulation.html …
\begin{thread}pic.twitter.com/vviJ1URnLd
The basic building blocks of SKY are:
1/ Consensus articles: @wikipedia like articles where scientists converge towards their interpretation of the evidence on a question.
2/ Continuously updated meta-analysis: the quantified evidence that supports the conclusionspic.twitter.com/Y0nEmMPuU6
3/ Dynamic research reports: containing both data and code, they can be reproduced and corrected. 4/ Methodological guidelines: agreed-upon evidence-based consensus on how to implement the empirical methods.pic.twitter.com/fLtJIMumv4
SKY will bring: - A collectively agreed-upon answer to the questions of the users of science. - A list of open questions for scientists and funders. - A focus on precision, aggregated evidence and on collective interpretation of the data - Storage for meta-analysis datapic.twitter.com/FwT99u0vda
Why SKY is needed? 1/ Social science needs unification of principles, results and methods 2/ There is no place to build such a consensus 3/ The current system locks knowledge behind paywalls, does not encourage collaboration and suffers from publication bias.pic.twitter.com/oJlOpQxzjr
What are the alternatives to SKY and why are they not enough?
1/ @Wikipedia is an inspiration, but has a No Original Research rule which prevents it from being a place to aggregate kwnowledgepic.twitter.com/vodiYob67z
2/ Several awesome initiatives conduct meta-analysis (@cochrane, @campbellreviews, Washington State Institute, @NICEcomms) but are not interactive, nor open source and they are sometimes paywalled.
3/ The Open Social Science initiatives (@OSFramework, @PsySciAcc) do not have as a goal to summarise literatures but they seek to conduct new studies.
4/ @JPAL_Global and @poverty_action have provided methodological guidelines and start to offer repos for datasets, but they do not provide aggregated knowledge.
5/ Twitter is great but not structured enough.
6/ Individual researchers such as @jenniferdoleac, myself and others have compiled literature reviews on google docs, but isn't it better to centralise them all in one place so that we can share and improve?
Why is SKY open source?
Science is a Bazaar, not a Cathedral.
Science is about collaboration not competition.
Science is self-corrective, and peer-review is key.
@wikipedia, @linuxfoundation show that such decentralized collaborative projects can succeed.pic.twitter.com/boOzq7mgQA
Which technology makes SKY possible?
- @MySQL and Google docs for storing data effectively.
- @GitHub for storing the code and the website and dealing with independent increments to the same project.
- Rmarkdown to generate a website combining R code and text.pic.twitter.com/qY1TB9KN3P
How to use and contribute to SKY? You can start using SKY now, by looking up the webpage and locating a question of interest for you and consulting what the answers are: https://chabefer.github.io/SKY/ You can contribute right now by posting me toward a meta-analysis missing on SKY.pic.twitter.com/bdbCVFtBKo
You can also download the project on GitHub and start making changes, adding pages and questions, editing the text an entry: https://github.com/chabefer/SKY You can find directions on how to use the main tools in SKY here: https://sites.google.com/site/sylvainchabeferret/research/git_mysql_knitr.pdf …
What can you contribute to SKY?
1/ Update meta-analysis results, or make me aware of them, or link to a preregistered one, or just ask a question. I'm following @EvidenceRobot but I still miss a LOT of meta-analysis. If you're aware of one, DM me or email me or update SKY.pic.twitter.com/GDTkZB5CEY
2/ Start an individual research report explaining how each result in a meta-analysis has been obtained. Link to the original paper and show how we go from the original paper to the entry in the meta-analysis. It's more than OK to link to research projects on other websites.pic.twitter.com/xyzzT2gwyj
3/ Contribute to the methodological guidelines. Make me aware of a new methodological paper, send me a link to google doc with a lit review. Put your simulations, results and code on SKY, so that people can test it.
Contribute to the open source book "Statistical Tools for Causal Inference" https://chabefer.github.io/STCI/
The Git repo is here: https://github.com/chabefer/STCI
I especially think of @causalinf and @nickchk here, but any econometrician or statistician is free to step up.
What are the rules of the game on SKY? - Be Bold - Do Not Strive for Perfection at First - Act in Good Faith - Assume Good Faith - Use Data to Settle Disputes - Do Not Proselityze - Do Not Violate Copyright.pic.twitter.com/uSU3WYNgtD
How is SKY funded?
Open Science is not costly (for the moment, bigger servers will be needed, I hope :).
I'm lucky enough to have employers that fully support open science: @Inra_France and @TSEinfopic.twitter.com/0NkpTNXEF3
What's in the books for SKY?
1/ Move to cloud-based computing to allow a full wikipedia-like feel. The current solution is technical and focuses on R.
Hey, @CodeOceanHQ, I'd love to work with you on this.
2/ Add a discussion chamber. For the moment, use a twitter thread :)pic.twitter.com/bxKtyMjTLk
3/ A voting system for referee remarks on individual papers, so that the most important ones get addressed first, like on @StackOverflow
4/ An anonymous mechanism for reporting abusive behavior.
5/ Private repos.
I'm dying to receive feedback about SKY and the plans for development.
Do respond in the thread, by DM or email.
I'll be @improvingpsych in Rotterdam this summer, really excited to hear what everyone has to say :)
A few HT:
@DinaPomeranz: this is me trying to smoothly disrupt scientific edition ;)
@simardcasanova: you were right, we need more meta-analysis in economics :)
@_alice_evans: thanks for the encouragement
@nalanpog: feel free to play with SKY :)
@chrisdc77: this is one way to achieve the vision in your book, hope you like it
@CRChartier: hopefully you'll find SKY useful for summarising results of @PsySciAcc
@BrianNosek: same for @osf results
@Treich13 and @alban_thomas_ts and @Xtophe_Bontemps: thanks for the support!
@TheReplicateNet: thanks so much for the encouragements :)
Next blog post will give you a live example of how things work on SKY :) \end{thread}
@jepusto and @RebeccaMaynar10: what are your thoughts? Does it achieve the vision you were putting forward?
@HeimArthur: you were the first to believe in SKY, thanks so much :)
HT to @BTCoachConseil: #SKY would not exist without your coaching
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
