At last week’s SSP conference in San Francisco, those of us interested in Altmetrics were rather excited to see representatives from each of the major products come together in a session entitled ‘Metrics 2.0: It’s about Time…..and People’.

First up was Andrea Michalek of Plum Analytics (who kindly shared her slides here). She revealed a sneak preview of work being done with the The Smithsonian, one of Plum’s first customers. Their product, PlumX, is being used to collect data (usage, captures, mentions, social media, citations) in order to generate reports on publication activity in support of research evaluation.

She explained how in scholarly communications, the same article can be published in multiple locations on the web (e.g. publisher website, PubMed Central, Mendeley). Fortunately, Plum collects and displays the counts from each of these individual locations, allowing users to get a full view of the engagement surrounding a particular article, video, presentation etc. Indeed, she stressed the importance of tracking the impact of all aspects of output, not just the article. She spoke of these ’2nd level metrics’ and used the example of an author who blogs about his/her research.

Next to speak was Euan Adie of Altmetric.com. He described his product as being more focused on publishers than researchers, which differentiates it from the other altmetric providers on the panel. He stressed that altmetrics are not identifiers of quality, the same way that citations are not. He went on to explain that articles that received the most online attention may well attract the ‘wrong’ audience and often contain populist terms like sex, coffee or cannabis in their title (more on this on the Altmetric.com blog).

Euan also presented a new feature that has been introduced around mentions in policy documents; clear indicators of impact that are not picked up by citations. While papers usually have 50% of the tweets they will ever receive in the first three days after publication, this is not so with policy documents. He presented an interesting example and asked where the evidence for the UK’s diabetes type 2 treatment guidelines comes from? Whilst the Lancet received 11% of the citations and the BMJ 9%, smaller journals also had an impact, such as CRMO, which received 2.6% (25 citations). Tracking this new metric will help us to surface the impact that smaller journals are having.

Jason Priem differentiated ImpactStory from the others by stressing that it was a non-profit organisation. Also, rather than focusing on traditional outputs, ImpactStory is interested in identifying alternative ways to measure and understand the new ecosystem of the Web. Just as the printing press transformed publishing, so too will the Web (to a larger degree than it has already). He explained that the Web is so powerful because there’s one bucket that contains everything. The  future of altmetrics should be the same and there needs to be an open infrastructure that can be built on top of (more on this later).

He raised an interesting issue about inconsistent results between different altmetric tracking systems. Why would a paper have different scores on altmetric.com and ImpactStory? Whilst he didn’t exactly answer the question, he did point to the fact that citation tracking services can also be wildly inconsistent, without causing too much inconvenience to researchers.

He showed an interesting example of how ImpactStory was being used. The UK Parkinson’s Disease Consortium at UCL have integrated ImpactStory into their publications page in order to show the impact of their output. This is especially important as they are trying to gain funding for further research.

Martin Fenner of PLOS was the final speaker. He reiterated that whilst there was a good deal of overlap between the various altmetric providers, there are significant differences (PLOS’ being that it is a publisher rather than a third party service). He ran through the various aspects of the PLOS article-level metric reports, including the ability to aggregate up statistics from specific sets of articles.  He stressed the importance of visualising data and showed an impressive 4D bubble chart showing total views, months, scopus ctitations and journal name.

Q&A

I’ve tried to capture some of the key points from the lively Q&A session that followed the main presentations:

• Although tweets may contain negative rather positive sentiment, altmetrics looks at a correlation of data points rather than just one metric. Also, it can sometimes be useful to see the negative impact.
• Authors really like to track their impact. However, this is not just vanity – they want feedback on their work.
• Researchers who have prospered in the traditional system of citations tend to be less enthusiastic about alternative metrics. Fans are often younger and more experimental users.
• Funders, such as the National Institutes of Health, are also enthusiasts as they need to show the value of the money they are spending.
• In terms of historical data, it’s very difficult to go back beyond a certain point. Twitter doesn’t even store data for more than 2 years. Altmetric.com goes back the furthest (to July 2011).

Towards the end of the session, a  representative from Emerald Group Publishing asked whether we will we need an aggregator to bring together all of the statistics provided by the (slightly confusing) array of altmetric services. Jason Priem’s response to this was really insightful. He expressed the desire for another party to build a ‘Google of altmetrics’ on top of all altmetric providers. He doesn’t see ImpactStory as being the end solution, more part of the infrastructure. It will certainly be interesting to see who and when somebody will step up to this challenge!

Since 1840, BMJ has been a trusted voice in the development of improved healthcare. We are proud of our heritage but also believe in looking forward. Our objective remains to support medical professionals and organisations in continuously improving the delivery of quality healthcare. By sharing our information, analytical tools and technology during an upcoming hack day (6-7 July), BMJ seeks to help healthcare professionals and organisations improve the care they provide.

What is a hack day?

A hack day (also known as a hackathon, hackfest or codefest) is an event in which computer programmers and others involved in software development, collaborate intensively on software projects. We’re working with Rewired State,  an organisation that runs hack days, and curates a network of more than 1,000 software developers and designers. Their clients include government bodies, large private companies, conferences and charities. They were behind National Hack the Government Day, currently in its fourth year. The first event was “the catalyst for the landmark site data.gov.uk and is widely acknowledged as the inspiration behind the UK government’s open data efforts.”

What are we looking for?

We’re inviting interested developers to focus on four categories:

1.   Digital Medical Students

Whether they’re in the lecture theatre or carrying out early rotations in A&E, what innovative applications could be built to create better doctors for our future healthcare needs?

2.   Create a ‘zero harm’ NHS

What could be done to prevent more unnecessary deaths, such as those seen at Mid Staff NHS Foundation Trust? There are 290 recommendations arising from the Francis report, which aims to create a ‘zero harm’ NHS.  Some recommendations have already been rejected or watered down in the government’s response. What innovative approaches can be used to kick start these recommendations, taking a radically different approach to traditional NHS technology procurement?

3.   Localise content through collaboration

Many developing countries look to westernised countries for the evidence on which to base their care. UK published materials, such as BMJ Best Practice, can provide healthcare professionals with a sound knowledge base but they need localisation to reflect not only differences in the prevalence of diseases but also in treatment options and how these treatments are delivered.

The challenge of localising content is significant. The cost of physical human translation is enormous and alongside the risk of mis­translating a key clinical concept or drug dose which could result in a loss of human life, to date it has not been considered commercially viable, if technically possible. Can crowdsourcing ever be applied to such a task, and if so, how would it work?

4.   Revolutionise the scholarly publishing process

The classical scholarly publishing model has not changed radically since the nineteenth century. A key criticism is the requirement to reduce years worth of research into a restrictive, text-based end product; the article of record. How could we capture more of the research process and enable collaboration before and after publication?

Logistics

Judges will also be announced soon.

This will be a two day hack, with overnight for those wanting it at Hub Westminster, London, on the weekend of 6th/7th July.

Intellectual Property Rights (IPRs) (subject to final confirmation): Content creators retain IPRs over original/hack day created code.

Code: Open-sourced with repositories publicly available encouraged.

Register for BMJ Hack Day >> 

Hack4ac – eLife

Also taking place on the Saturday is Hack4ac, which presents a great opportunity for collaboration over the two hack events. Hack4ac will focus on the journal and content side, and the BMJ hack day is going to focus on medical data and hacks that could help with front line patient support.

They have two goals:

• Demonstrate the value of the CC-BY licence within academia. We are interested in supporting innovations around and on top of the literature.
• Reach out to academics who are keen to learn or improve their programming skills to better their research. We’re especially interested in academics who have never coded before.

We are going to coordinate on the weekend of the event, and in the run-up to see if we can cross-fertilise ideas. We will be announcing a pub event in the run-up to the weekend so register above to receive updates.

This week has seen a flurry of activity on Twitter owing to a series of separate but related events highlighting trends in scholarly communication and research data. The fun kicked off with Wednesday’s ‘Now and Future of Data Publishing’ event organised by the Jisc Managing Research Data programme. It was followed on Thursday by the  ORCID Outreach Meeting, Getting Credit for Your Work: A Symposium on Research Attribution (jointly organised by Dryad and ORCID) and concludes with today’s Dryad Membership Meeting.

An obvious but important message that underpinned discussions on all three days was the importance of sharing data.  On the first morning, Simon Hodson of Jisc quoted Geoffrey Boulton of the Royal Society (who have made sharing data a condition of publication): “Publishing articles without making the data available is scientific malpractice.” This is an extreme but not uncommon view.

Trish Groves, deputy editor of the BMJ, recently wrote a summary of recent and future developments around sharing clinical data. The big news our end is that we now require authors of drug and devices trials to deposit their anonymised patient level data—on reasonable request. However, our interest in data publication started back in 2009 when we first incorporated data sharing statements into all BMJ research papers. More recently, we have encouraged our authors to deposit their data in Dryad and the first article to link through to a Dryad dataset was published by BMJ Open in 2011. We now have just under 40 papers with links to datasets on Dryad.

BMJ is by no means the only publisher experimenting with how best to integrate associated research data into published articles. While we link through to individual data sets on Dryad from our data sharing statements, there are a number of other options being developed.

I first blogged about FigShare back in Feb 2012. Since then, they’ve collaborated with a number of big publishers, including PLOS, Faculty of 1000 and Nature Publishing Group. They offer a distinctly unique set of embeddable widgets, which allow readers to view the underlying data within the article view rather than clicking through to a separate website (see an example from PLOS below).

FigShare is also collaborating with Nature Publishing Group (NPG) on Scientific Data, one of a new breed of data journals for articles describing datasets deposited elsewhere, including biomedical research data. Notably, it will not be a condition of publication that any papers referring to the datasets be published in NPG journals.

As more funders and learned societies call for new ways to make research data more available, reusable and reproducible, it will be interesting to see how established and emerging platforms will work with researchers and publishers to make access to data as pain-free as possible.

In the past month, almost 1 million new magazines have appeared on Apple’s iPad. Rather than heralding a long-awaited comeback from traditional publishers, nearly all of these collections of articles, photos and social-media updates are the handiwork of ‘armchair editors’, using a new tool from the social magazine app, Flipboard.

more…

Research and discovery in the life sciences is a pretty complicated business. The complexity of the modern scientific process seems to be a reflection of the intricacies of life and the processes associated with disease and its treatment. Furthermore, as technologies become more advanced, so too does the problem of managing the ever expanding quantity of data being generated.

Currently, pharmaceutical companies expend significant and duplicated efforts aligning and integrating their internal information with public data sources. This process is largely incompatible with large-scale computational approaches and the vast majority of drug discovery sources find it difficult to complicate with eachother. more…

This week, Nature Publishing Group announced they had reached 35,000 followers on Pinterest. Given that they had just 127 followers at the end of 2012, this is phenomenal growth. How did they achieve it, I hear you ask? When asked the secret to their success on Twitter, @NatureBlogs replied: “best tip is to continuously update the boards with new images. We make daily (if we can) updates”.  So, given the effort and resource needed to maintain this level of engagement, what can publishers expect in return? Do follower numbers translate into traffic to journal websites?

Over a year ago, I wrote an introductory blog covering the basic mechanics of Pinterest and the impressive growth it had experienced:  it is still the fastest standalone website to surpass the 10 million mark. Now with more than 48 million users, Pinterest is one of the most popular social networks on the web. So, how can publishers get a slice of the action?

Add keywords

Pinterest is a popular site but it’s not nearly as saturated as Google search results. While it’s true that most people prefer to browse Pinterest than search it, there are a significant number who want to discover something that hasn’t already been repinned many times, and search is where they do it.

With every image you post, you should include a clear description that people will enjoy reading. The key seems to be mentioning a keyword that reveals few results in Pinterest but is likely to be searched for often. This can help get the exposure necessary to be repinned and therefore reach more people.

Similarly, you can also take advantage of Pinterest’s popularity to get your Pinterest page into Google. Link to your pinboard from your website, during online promotions and from established social media channels to improve the chances of it showing up in Google search results.

Repin others

In a similar vein to Twitter’s retweets, build awareness of your Pinterest account by repinning others. Use both the search function and categories to find other pins relevant to your boards. When you repin another’s pinned picture, they will receive an email notification. Hopefully, they will follow the links to your account and see where you pinned their picture, giving you an opportunity to have them look through your boards. The obvious need here is to have boards that interest your new visitor to the point that they consider following your links to the site you are promoting.

Comment on pins

When you hold your cursor over a pinned picture you see the “comment” tab, the “repin” tab, and the “like” tab. When you make a comment is stays with the picture, so whatever you say has the potential to catch people’s attention and drive traffic. You might even consider some kind of “call to action” (if it’s appropriate). Again, Pinterest will notify the person who pinned the picture that a comment has been made and give them a link to respond to your comment.

Check your stats

Pinterest introduced a web-based analytics tool in March, allowing site owners to track users’ engagement with their sites on the social network.

The free tool allows site owners to track the number of pinners and pins collecting material from their sites, and the number of repinners and repins those initial pins received. Site owners can also track total impressions and reach on the network, as well as referral traffic, both in clicks and unique visitors, sent back to their sites. This is key in understanding your audience’s likes/dislikes, which will feed into the future selection of content to add.

Publishers with successful Pinterest profiles

• Nature Publishing Group
• Forbes Magazine
• The Guardian
• Time Magazine
• The New York Times
• Wall Street Journal
• BBC

In last week’s blog I looked at the innovative ways that publisher’s are monetising their products in the face of a shifting digital landscape. One of the most ground-breaking moves has come from Forbes Media, who opened up their content creation platform, not only to external authors but also to marketers and brands.

This idea of interweaving promotional content with both editorial and user-generated content (UGC) is central to a much larger concept gaining traction in the online advertising community; native advertising.

There were a number of key themes at this year’s Guardian Changing Media Summit — an annual conference which brings together a mixture of CEO and director level executives responsible for commercial, creative and digital strategies.

Most significantly, it was clear that publishers are beginning to see more opportunities than threats from digital technologies and much time was spent discussing the innovative monetisation of digital products.

more…

As you may already be aware, Facebook is rolling out the first major update to its News Feed since the feature launched nearly seven years ago. As with every other change the site has made, the new design has been met with mixed reactions and hasn’t gone unnoticed by the media.

Facebook’s revamped News Feed gives the homepage a starkly mobile look, reducing clutter and lending more space to prominent photographs. It takes significant cues from the Facebook mobile apps for phones and tablets, adding a new side navigation bar and more white space.

more…

If you’re looking to connect with people in a particular niche on Twitter, Followerwonk could be just the tool for you. It’s currently free to search Twitter biographies, compare users and analyse followers of multiple accounts, so try it out before subscriptions kick in.

What can I find out about my followers?

By linking a Twitter account to Followerwonk, users can run a number of different analytic reports for free. Below is a list of the most useful for strategically growing a following and connecting with ‘influencers’ in a specific area:

• Influence scores – how influential are your followers?
• Follower counts – how many followers do your followers have?
• Mapped locations – where are your followers located? (see below) more…