Improving openness, transparency and reproducibility in scientific research

Jenny Molloy - October 24, 2014 in Guest post, Reproducibility, research, tools

This is a guest post by Sara Bowman of the Open Science Framework.

Understanding reproducibility in science

Reproducibility is fundamental to the advancement of science. Unless experiments and findings in the literature can be reproduced by others in the field, the improvement of scientific theory is hindered. Scholarly publications disseminate scientific findings, and the process of peer review ensures that methods and findings are scrutinized prior to publication. Yet, recent reports indicate that many published findings cannot be reproduced. Across domains, from organic chemistry ((Trevor Laird, “Editorial Reproducibility of Results” Organic Process Research and Development) to drug discovery (Asher Mullard, “Reliability of New Drug Target Claims Called Into Question” Nature Reviews Drug Development) to psychology (Meyer and Chabris, “Why Psychologists’ Food Fight Matters” Slate), scientists are discovering difficulties in replicating published results.

Various groups have tried to uncover why results are unreliable or what characteristics make studies less reproducible (see John Ioannidis’s “Why Most Published Research Findings Are False,” PLoS, for example). Still others look for ways to incentivize practices that promote accuracy in scientific publishing (see Nosek, Spies, and Motyl, “Scientific Utopia II: Restructuring Incentives and Practices to Promote Truth Over Publishability” Perspectives on Psychological Science). In all of these, the underlying theme is the need for transparency surrounding the research process – in order to learn more about what makes research reproducible, we must know more about how the research was conducted and how the analyses were performed. Data, code, and materials sharing can shed light on research design and analysis decisions that lead to reproducibility. Enabling and incentivizing these practices is the goal of The Open Science Framework, a free, open source web application built by the Center for Open Science.

The right tools for the

The Open Science Framework (OSF) helps researchers manage their research workflow and enables data and materials sharing both with collaborators and with the public. The philosophy behind the OSF is to meet researchers where they are, while providing an easy means for opening up their research if it’s desired or the time is right. Any project hosted on the OSF is private to collaborators by default, but making the materials open to the public is accomplished with a simple click of a button.

Here, the project page for the Reproducibility Project: Cancer Biology demonstrates the many features of the Open Science Framework (OSF). Managing contributors, uploading files, keeping track of progress and providing context on a wiki, and accessing view and download statistics are all available through the project page.

Here, the project page for the Reproducibility Project: Cancer Biology demonstrates the many features of the Open Science Framework (OSF). Managing contributors, uploading files, keeping track of progress and providing context on a wiki, and accessing view and download statistics are all available through the project page.

Features of the OSF facilitate transparency and good scientific practice with minimal burden on the researcher. The OSF logs all actions by contributors and maintains full version control. Every time a new version of a file is uploaded to the OSF, the previous versions are maintained so that a user can always go back to an old revision. The OSF performs logging and maintains version control without the researcher ever having to think about it – no added steps to the workflow, no extra record-keeping to deal with.

The OSF integrates with other services (e.g., GitHub, Dataverse, and Dropbox) so that researchers continue to use the tools that are practical, helpful, and a part of the workflow, but gain value from the other features the OSF offers. An added benefit is in seeing materials from a variety of services next to each other – code on GitHub and files on Dropbox or AmazonS3 appear next to each other on the OSF – streamlining research and analysis processes and improving workflows.

 Each project, file, and user on the OSF has a persistent URL, making content citable. The project in this screenshot can be found at https://osf.io/tvyxz.

Each project, file, and user on the OSF has a persistent URL, making content citable. The project in this screenshot can be found at https://osf.io/tvyxz.

Other features of the OSF incentivize researchers to open up their data and materials. Each project, file, and user is given a globally unique identifier – making all materials citable and ensuring researchers get credit for their work. Once materials are publicly available, the authors can access statistics detailing the number of views and downloads of their materials, as well as geographic information about viewers. Additionally, the OSF applies the idea of “forks,” commonly used in open source software development, to scientific research. A user can create a fork of another project, to indicate that the new work builds on the forked project or was inspired by the forked project. A fork serves as a functional citation; as the network of forks grows, the interconnectedness of a body of research becomes apparent.

Openness and transparency about the scientific process informs the development of best practices for reproducible research. The OSF seeks both to enable that transparency, by taking care of “behind the scenes” logging and versioning without added burden on the researcher – and to improve overall efficiency for researchers and their daily workflows. By providing tools for researchers to easily adopt more open practices, the Center for Open Science and the OSF seek to improve openness, transparency, and – ultimately – reproducibility in scientific research.

Global Open Data Index 2014 – Week ending October 24: Update

Mor Rubinstein - October 24, 2014 in Open Data Index

Skærmbillede 2014-10-23 kl. 23.41.52

Thank you so much for the amazing number of submissions we have received this week.

Entering the final week of the global sprint!

Next week is the FINAL WEEK of the global Index sprint. Please make sure your country is represented in the Global Open Data Index 2014.

Find out how you can contribute here.

If you would like to be reviewer for Global Open Data Index 2014, please sign up here.

We are missing some countries – can you help?

Europe – Armenia, Croatia, Hungary, Ukraine, Slovenia, Norway, Finland, Denmark, Sweden, Portugal, Poland

Americas – Honduras, Nicaragua, Guatemala, Brazil, USA

Asia/Pacific – South Korea, Taiwan, India, Mongolia, New Zealand, Australia

Africa – Sierra Leone, Mali, Malawi, Mozambique 

Join the Global Madness Day: October 30

Take part in a day of activities to make sure we get the most submissions through that we can for the Global Open Data Index 2014. Make sure your country is represented – the October 30 Global Madness Day is the last day in the sprint!

At 2pm GMT Rufus Pollock, the President & Founder of Open Knowledge will be chatting to Mor Rubinstein about the Index in a Google Hangout. Make sure you join the chat here!

Other events will take place throughout the day. See our twitter feed for updates #openindex14

Some practical tips…

Lastly, a couple of reminders on some key questions around the Index from Mor Rubinstein, Community coordinator for the Index:

1. What is machine readable? – This year we added help text for this question. Please read it when making your submissions this year. Frequently contributors categorise HTML format as a machine readable format. While it is easy to scrape HTML, it is actually NOT a machine readable format. Please use our guide if you are in doubt or send an email to the census list.

2. What is Openly Licensed? – Well, most of us are not lawyers, and the majority of us never pay attention to the term and conditions on a website (well, they are super long… so I can’t blame any of you for that). If you are confused, go to the Open definition which gives a one page overview on the subject.

Open Access and the humanities: on our travels round the UK

Guest - October 23, 2014 in Open Access, Open Access Week

This post is part of our Open Access Week blog series to highlight great work in Open Access communities around the world. It is written by Alma Swan, Director of Key Perspectives Ltd, Director of Advocacy forSPARC Europe, and Convenor for Enabling Open Scholarship.

Digital Camera

Whither the humanities in a world moving inexorably to open values in research? There has been much discussion and debate on this issue of late. It has tended to focus on two matters – the sustainability of humanities journals and the problem(s) of the monograph. Neither of these things is a novel topic for consideration or discussion, but nor have solutions been found that are satisfactory to all the key stakeholders, so the debate goes on.

While it does, some significant developments have been happening, not behind the scenes as such but in a quiet way nevertheless. New publishers are emerging in the humanities that are offering different ways of doing things and demonstrating that Open Access and the humanities are not mutually exclusive.

These publishers are scholar-led or are academy-based (university presses or similar). Their mission is to offer dissemination channels that are Open, viable and sustainable. They don’t frighten the horses in terms of trying to change too much, too fast: they have left the traditional models of peer review practice and the traditional shape and form of outputs in place. But they are quietly and competently providing Open Access to humanities research. What’s more, they understand the concerns, fears and some bewilderment of humanities scholars trying to sort out what the imperative for Open Access means to them and how to go about playing their part. They understand because they are of and from the humanities community themselves.

The debate about OA within this community has been particularly vociferous in the UK in the wake of the contentious Finch Report and the policy of the UK’s Research Councils. Fortuitously, the UK is blessed with some great innovators in the humanities, and many of the new publishing operations are also UK-based. This offers a great opportunity to show off these some new initiatives and help to reassure UK humanities authors at the same time. So SPARC Europe, with funding support from the Open Society Foundations, is now endeavouring to bring these new publishers together with members of the UK’s humanities community.

We are hosting a Roadshow comprising six separate events in different cities round England and Scotland. At each event there are short presentations by representatives of the new publishers and from a humanities scholar who can give the research practitioner perspective on Open Access. After the presentations, the publishers are available in a small exhibition area to display their publications and talk about their publishing programmes, their business models and their plans for the future.

The publishers taking part in the Roadshow are Open Book Publishers, Open Library of the Humanities, Open Humanities Press and Ubiquity Press. In addition, the two innovative initiatives OAPEN and Knowledge Unlatched are also participating. The stories from these organisations are interesting and compelling, and present a new vision of the future of publishing in the humanities.

Humanities scholars from all higher education institutions in the locality of each event are warmly invited to come along to the local Roadshow session. The cities we are visiting are Leeds, Manchester, London, Coventry, Glasgow and St Andrews. The full programme is available here.

We will assess the impact of these events and may send the Roadshow out again to new venues next year if they prove to be successful. If you cannot attend but would like further information on the publishing programmes described here, or would like to suggest other venues the Roadshow might visit, please contact me at sparceurope@arl.org

Harvard Library lifts copyright restrictions on public domain works

Lieke Ploeger - October 23, 2014 in Featured, News

As part of the international Open Access Week (20-26 October), Harvard Library shared great news on their new policy on the use of digital reproductions of public domain works. From now on, the library will make such reproductions openly available online and treat them as objects in the public domain. This means that users will be able […]

New Open Access Button launches as part of Open Access Week

David Carroll - October 22, 2014 in Featured Project, Open Access, Open Access Button

This post is part of our Open Access Week blog series to highlight great work in Open Access communities around the world.

button

Push Button. Get Research. Make Progress.

If you are reading this, I’m guessing that you too are a student, researcher, innovator, an everyday citizen with questions to answer, or just a friend to Open Knowledge. You may be doing incredible work and are writing a manuscript or presentation, or just have a burning desire to know everything about anything. In this case I know that you are also denied access to the research you need, not least because of paywalls blocking access to the knowledge you seek. This happens to me too, all the time, but we can do better. This is why we started the Open Access Button, for all the people around the world who deserve to see and use more research results than they can today.

Yesterday we released the new Open Access Button at a launch event in London, which you can download from openaccessbutton.org. The next time you’re asked to pay to access academic research. Push the Open Access Button on your phone or on the web. The Open Access Button will search the web for version of the paper that you can access.

If you get your research, you can make progress with your work. If you don’t get your research, your story will be used to help change the publishing system so it doesn’t happen again. The tool seeks to help users get the research they need immediately, or adds papers unavailable to a wish-list we can get started . The apps work by harnessing the power of search engines, research repositories, automatic contact with authors, and other strategies to track down the papers that are available and present them to the user – even if they are using a mobile device.

The London launch led other events showcasing the Open Access Button throughout the week, in Europe, Asia and the Middle East. Notably, the new Open Access Button was previewed at the World Bank Headquarters in Washington D.C. as part of the International Open Access Week kickoff event. During the launch yesterday, we reached at least 1.3 million people on social media alone. The new apps build upon a successful beta released last November that attracted thousands of users from across the world and drew lots of media attention. These could not have been built without a dedicated volunteer team of students and young researchers, and the invaluable help of a borderless community responsible for designing, building and funding the development.

Alongside supporting users, we have will start using the data and the stories collected by the Button to help make the changes required to really solve this issue. We’ll be running campaigns and supporting grassroots advocates with this at openaccessbutton.org/action as well as building a dedicated data platform for advocates to use our data . If you go there you now you can see the ready to be filled map, and your first action, sign our first petition, this petition in support of Diego Gomez, a student who faces 8 years in prison and a huge monetary fine for doing something citizens do everyday, sharing research online for those who cannot access it.

If you too want to contribute to these goals and advance your research, these are exciting opportunities to make a difference. So install the Open Access Button (it’s quick and easy!), give it a push, click or tap when you’re denied access to research, and let’s work together to fix this problem. The Open Access Button is available now at openaccessbutton.org.

Building Community Action at Mozilla Festival 2014

Beatrice Martini - October 22, 2014 in #mozfest, community

Often Community is thought of as a soft topic. In reality, being part of a community (or more!) is admirable, a wonderful effort, both very fun but also sometimes tough and building and mobilising community action requires expertise and understanding of both tools and crowds – all relationships between stakeholders involved need to be planned with inclusivity and sustainability in mind.

10481478763_12fd0db3a2_o

This year Mozilla Festival (London, October 24-26), an event we always find very inspiring to collaborate with, will feature a track focusing on all this and more. Called Community Building, and co-wrangled by me and Bekka Kahn (P2PU / Open Coalition), the track has the ambitious aim to tell the story about this powerful and groundbreaking system, create the space where both newcomers and experienced community members can meet, share knowledge, learn from each other, get inspired and leave the festival feeling empowered and equipped with a plan for their next action, of any size and shape, to fuel the values they believe in.

We believe that collaboration between communities is what can really fuel the future of the Open Web movement and we put this belief into practice from our curatorship structure (we come from different organisations and are loving the chance to work together closely for the occasion) to the planning of the track’s programme, which is a combination of great ideas that were sent through the festival’s Call for Proposals and invitations we made to folks we knew would have had the ability to blow people’s mind with 60 minutes and a box of paper and markers at their disposal.

The track has two narrative arcs, connecting all its elements: one focusing on the topics which will be unpacked by each session, from gathering to organising and mobilising community power and one aiming to embrace all learnings from the track to empower us all, member of communities, to take action for change.

The track will feature participatory sessions (there’s no projector is sight!), an ongoing wall-space action and a handbook writing sprint. In addition to this, some wonderful allies, Webmaker Mentors, Mozilla Reps and the Space Wranglers team will help us make a question resonate all around the festival during the whole weekend: “What’s the next action, of any kind/ size/ location, you plan to take for the Open Web movement?”. Participants to our track, passer-bys feeding our wall action, folks talking with our allies will be encouraged to think about the answer to this, and, if not before, join our space for our Closing Circle on Sunday afternoon when we’ll all share with each other our plans for the next step, local or global, online or offline, that we want to take.

Furthermore, we also invite folks who’ll not be able to join us at the event to get in touch with us, know more about what we’re making and collaborate with us if they wish. Events can be an exclusive affair (they require time and funds to be attended) and we want to try to overcome this obstacle. Anyone will be welcome to connect with us in (at least) three ways. We’ll have a dedicated hashtag to keep all online/remote Community conversations going: follow and engage with #MozFestCB on your social media platform of choice, we’ll record a curated version of the feed on our Storify. We’ll also collect all notes, resources of documentation of anything that will happen in and around the track on our online home. The work to create a much awaited Community Building Handbook will be kicked off at MozFest and anyone who thinks could enrich it with useful learnings is invited to join the writing effort, from anywhere in the world.

If you’d like to get a head start on MozFest this year and spend some time with other open knowledge community minded folks, please join our community meetup on Friday evening in London.

Launching a new book on Responsible Development Data

Zara Rahman - October 22, 2014 in Open Development

RDF-book-cover-724x1024

This week I, and a group of development data experts from around the world, met for three days in a small farmhouse in the Netherlands to produce a book on Responsible Development Data. Today, we’re very happy to launch the first version: comments and feedback are really welcome, and please feel free to share, remix, and re-use the content.

Download the book here: http://tiny.cc/rddevbook

This book is offered as a first attempt to understand what responsible data means in the context of international development programming. We have taken a broad view of development, opting not to be prescriptive about who the perfect “target audience” for this effort is within the space. We also anticipate that some of the methods and lessons here may have resonance for related fields and practitioners.

The group of contributors working on this book brings together decades of experience in the sector of international development; our first hand experiences of horrific misuse of data within the sector, combined with anecdotal stories of (mis)treatment and usage of data having catastrophic effects within some of the world’s most vulnerable communities, has highlighted for us the need for a book tackling issues of how we can all deal with data in a responsible and respectful way.

Why this book?

It might have been an uneasy sense that the hype about a data revolution is overlooking both the rights of the people we’re seeking to help and the potential for harm that accompanies data and technology in development context. The authors of this book believe that responsibility and ethics are integral to the handling of development data, and that as we continue to use data in new, powerful and innovative ways, we have a moral obligation to do so responsibly and without causing or facilitating harm. At the same time, we are keenly aware that actually implementing responsible data practices involves navigating a very complex, and fast-evolving, minefield – one that most practitioners, fieldworkers, project designers and technologists have little expertise on. Yet.

We could have written another white paper that only we would read, or organised another conference that people would forget about. We tried instead to pool our collective expertise and concerns, to produce a practical guide that would help our peers and the wider development community to think through these issues. With the support of Hivos, Book Sprints and the engine room, this book was collaboratively produced (in the Bietenhaven farm, 40 minutes outside of Amsterdam) in just three days.

The team: Kristin Antin (engine room), Rory Byrne (Security First), Tin Geber (the engine room), Sacha van Geffen (Greenhost), Julia Hoffmann (Hivos), Malavika Jayaram (Berkman Center for Internet & Society, Harvard), Maliha Khan (Oxfam US), Tania Lee (International Rescue Committee), Zara Rahman (Open Knowledge), Crystal Simeoni (Hivos), Friedhelm Weinberg (Huridocs), Christopher Wilson (the engine room), facilitated by Barbara Rühling of Book Sprints.

Mission Statement der deutschsprachigen Open Science AG veröffentlicht

Christian Heise - October 21, 2014 in Deutschland, Featured, offenes Wissen, Open Access, Open Science

Zur internationalen Open Access Week 2014 hat die AG Open Science ein Mission Statement veröffentlicht. Ziel der Arbeitsgruppe ist die Vernetzung von Aktiven im Bereich Öffnung von Wissenschaft und Forschung (Open Science) und die Erarbeitung rechtssicherer Rahmenbedingungen für das Veröffentlichen von Forschungsergebnissen. Zusätzlich soll die Arbeitsgruppe die Zusammenarbeit mit anderen internationalen Open Science Gruppen koordinieren und als Ansprechpartner für Forscher, Institute, Zivilgesellschaft, Wirtschaft und Politik zum Thema Open Science fungieren.

Die Open Science AG ist eine offene Initiative, welche die Idee freier und offener Wissenschaft und Forschung in Wirtschaft, Gesellschaft und Politik etablieren möchte. Ihre Mitglieder rekrutieren sich aus verschiedenen Teilen der Forschungslandschaft und assozierter Gebieten. Sie sind überzeugt, dass offen praktizierter Forschung eine Verbesserung der Effizienz der Wissenschaft und langfristig positive Effekte für die gesamte Gesellschaft bewirken wird. Die Öffnung der Wissenschaft soll auf verschiendenen Ebene vorangetrieben werden. So unterstützt die Open Science AG WissenschaftlerInnen bei der Arbeit mit digitalen Ressourcen und Methoden in Forschung und Lehre. Sie will Kooperationsweisen, Methoden und digitale Werkzeuge optimieren, so dass Open Science ohne erheblichen Mehraufwand realisiert werden kann. Dazu gehört ebenfalls, dass WissenschaftlerInnen entsprechende Anerkennung wie auch finanzielle Mittel für die Öffnung der eigenen Forschung erhalten. Auch die Einbeziehung von Bürgern in den Wissenschaftsprozess (Citizen Science) soll vorangetrieben werden. Die AG lebt von den Ideen und der Iniative Ihrer Mitglieder und lädt daher alle Interessiert ein sich einzubrigen.

Sie wollen mehr erfahren oder gleich aktiv werden? Mehr über die Arbeit der AG, ihr Mission Statment sowie über die Mitglieder findet man ab sofort unter http://okfn.de/open-science/ oder Sie abonnieren die Mailingliste der Open Science AG.

New Broken Link Checker Plugin for CKAN

Sean Hammond - October 21, 2014 in Extensions, Presentations

deadoralive is a new broken-link checker service that works with CKAN and other sites, and ckanext-deadoralive is a CKAN plugin that you can install to integrate your CKAN site with the link checker. The pair have been developed by Open Knowledge as part of our work on the new version of Öppnadata, the Swedish national open data portal.

This quick presentation gives an overview of the link checker’s features and design:

For more details, see these blog posts:

Veröffentlichung Open Definition 2.0 & deutsche Übersetzung

Christian Heise - October 21, 2014 in Featured, offene Daten, Open Definition, Open Knowledge Foundation

open-definition-deBereits vor knapp zwei Wochen wurde die zweite Version der Open Definition veröffentlicht. Sie definiert Grundprinzipien für Open Data, Open Content und schreibt unsere Grundprinzipien für Offenheit bei Open Data und offenen Inhalten aller Art jenseits von Open-Source-Software fest. Seit 7. Oktober ist sie in der überarbeiteten Version 2.0 verfügbar.

Ziel der Open Definition war und ist es OpenWashing identifizierbar zu machen und zu vermeiden, sowie rechtliche Probleme auf Grund der Verwendung falscher oder nicht kompatibler Open-Lizenzen präventiv zu verhindern. Ebenfalls neu: die Daten-Deutschland-Lizenz 2.0 wurde in diesem Zusammenhang als kompatible Lizenz mit aufgenommen (wie bereits angekündigt).

Auch wenn es bereits Kritikpunkte an der neuen Version der Defintition gibt, rufen Christian Hauschke und Adrian Pohl dazu auf bei der Übersetzung der Version 2.0 in die deutsche Sprache zu helfen. Dabei sollen eventuellen Schwächen in der Definition nicht Aufgabe der Übersetzung sein. Wir unterstützen diesen Aufruf. Bei der Übersetzung kann im Pad unter https://pad.okfn.org/p/OpenDefinitionde geholfen werden.