Monthly Archives: December 2016

November Downloads in ORO

This months Top 50 includes one of the oldest records in ORO at number 40.  The eprint number is 38 (we are now on eprint number 48065 and counting), which indicates it was one of the first to be added to the ORO eprints software.

The paper  Ekins, Paul; Simon, Sandrine; Deutsch, Lisa; Folke, Carl and De Groot, Rudolf (2003). A Framework for the practical application of the concepts of critical natural capital and strong sustainability. Ecological Economics, 44(2-3) pp. 165–185 was deposited over 10 years ago in ORO and according to the eprints software has had 3,762 downloads and 422 views.  Over half of those downloads have come direct from Google (or Google Scholar).  

The full text of the paper is also in ResearchGate which records 137 reads – although I’m not entirely sure when it was added to ResearchGate or exactly what a “Read” is!  It is also in Repec which is a subject repository for Economics – but there is no full text archived there. The full text is also in CORE and it looks like it was a pretty early addition to CORE if the url https://core.ac.uk/display/8 is anything to go by.

Downloads of the item in ORO have grown over the years in a not dissimilar pattern to how a paper accrues citations:

38downloads

Citation data in Scopus indicate that this is a very highly cited paper, accruing over 240 citations, with a Field-Weighted Citation Impact of 4.34.  In terms of citations it appears to be in the top 2% of papers, based on its subject area, date and type of publication.  Citation pattern from Scopus is below:

scopus citations

I’m not attempting to make any correlation between downloads from ORO and citations from this single instance… I have made some exploratory analysis before.  It’s just rewarding that something archived in the infancy of the University repository is a highly cited paper, was originally made Open Access via ORO and remains a popular paper in terms of downloads today.

Top November downloads from ORO are below:

November Downloads

Author Profile Systems

Author Profile Systems (like Academia.edu, ResearchGate, ResearcherID and Google Scholar profiles) have grown in number over the last few years. The prevalence of them begs the question as to which one(s) to use, or, should you use any at all!  So in this post I’ll try and outline some of the benefits of the systems, their different characteristics, and some particular things you should consider when creating and curating author profile systems.

The benefits

Simply put, the benefits of Author Profile Systems are:

  • To make your research known
  • To increase chance of citation
  • To correct attributions
  • To ensure research is counted in research assessment
  • To increase chance of new collaboration
  • To increase chance of funding

The differing characteristics of the systems

Different functions have created different Author Profile Systems – that might sound awkward and I intend it to be.  What I mean is that the primary function of a system (e.g. author disambiguation, open access, reference management or search engines) has created different types of profile systems.  So the characteristics of a Google Scholar profile will in part be defined by the original function of Google Scholar, that being a search engine.  However, the common feature of all profile systems is the attempt to increase the visibility of the researcher and their work.

Copy of Disambiguation

A further point to consider is that different Author Profile Systems have different modus operandi which will affect your interaction with them.  Some are institutional, others are personal; some commercial, others not for profit; some are open, others are closed (or should I say are a walled garden).  The tension between these aspects may determine how you interact (or choose not to interact) with any particular profile.  I have written before on the considerations anyone should make when using commercial profile systems like ResearchGate and Academia.edu.

Readymade Systems

Importantly, some profiles already exist, without you having to create them yourself. These include ScopusID (the Elsevier persistent identifier for researchers) and OU People Profile pages.  For ScopusID this is important as Scopus attempts to match publications to an existing author profile or create a new author profile – this doesn’t always work, it’s an algorithm.  So in Scopus you can get multiple author profiles for the same person and you can get papers associated to a particular author’s profile that weren’t written by them at all.  So this profile needs to be curated.

For the OU Profile pages the problem is different, but the solution remains curation.  The profile pages are automatically created so if you don’t add any information to them then they are basically empty – a zombie profile, so to speak.  And, unfortunately if you own a zombie profile and do a google search on your name and “Open University” that zombie page is going to appear pretty close to the top of a search results page.  Which I don’t think is desirable.

So here is a bit of checklist of author profile systems and what they ostensibly can and cannot do.  There are a couple of caveats to this checklist.   Firstly, functionality changes very quickly on some of these systems so what it can’t do one week it may be able to do the next.  Secondly, not all Yes’s are equal, for instance whilst ResearchGate and ORO can provide metrics, it is a qualitatively different type of metrics to that offered by ResearcherID or ScopusID.

APAuthor Profile Systems PDF

What should you do?

Well firstly you should look after the OU People Profile and the ORO publications that populate the publications page.  There are external drivers (i.e. REF) to maintain a publication record in ORO and there are significant benefits in making publications Open Access in ORO wherever possible.  Add some biography to the People Profile and keep it current.

Secondly, look after any other profile that is readymade.  This at the moment might only be a light touch review of your Scopus account – this shouldn’t be onerous, but it is important as it may be affecting author level metrics created in Scopus.

Thirdly, get an ORCID.  I’ve written about the benefits of ORCID before.  ORCIDs are quickly becoming the de facto standard in scholarly communications because they are open and non-proprietary.  They are increasingly expected (if not necessarily required) by publishers and funders.

And, then, I’d suggest, take your pick based on any key criteria you have.  If you are an Early Career Researcher attempting to make connections for collaboration and funding then using a commercial platform to help you seems a sensible option.  Choose a platform you are comfortable with and where your peers have a presence.  However, if you already have those connections you may not need to take the trouble.

 

Metrics: what they are and how to use them

David and I ran the session on metrics on Wednesday and had a good attendance – so thanks to all who came out.

metrics-what they are and how to use them-December-2016- Blog

The session was an overview of metrics, researcher profiles, the application of metrics and the problematic behaviours these applications have created.

The slides are available: metrics-what they are and how to use them-December-2016- Blog and so are our notes:  Metrics-what-they-are-and-how-to-use-them-script-December-2016.

 

 

Where do Institutional Repository (ORO) Downloads come from?

Checking monthly download statistics gives a great insight into individual cases of how events in social media can impact the dissemination of Open Access research outputs. In October the article An investigation into written comments on assignments: do students find them usable? by Mirabelle Walker in Assessment & Evaluation in Higher Education, 34(1) received an exceptional 531 downloads.  But unfortunately I couldn’t track down the event(s) that caused this spike.  All I could see was a large set of referrals from a Facebook post (which I couldn’t get any more specific about…. gggrrrr!!!) and also a large set of referrals from Greece!

Top downloads for both September and October are below.

September Top Downloads

October Top Downloads

So I stepped back and decided to think about the bigger picture – how do items get downloaded from ORO i.e. what platforms are being used to download items from ORO and exactly where in the world do they come from.

Firstly, I compared the number of external referrals to referrals from ORO itself (e.g. someone clicking the download file from an external website rather than downloading an item from within ORO itself).

InternalExternalReferrals

Internal referrals equalled 604,344 (44%), whereas external referrals equalled 783,232 (56%)… most users of ORO never actually visit the site!

So from here I looked at the most popular external referrers (no prizes are on offer…)

Audience referrrers_v.2Yes, ORO loves Google (82.69%, 642,185 downloads).  Unpicking it a bit I was surprised how little traffic gets referred from Twitter (0.19%, 1,438 downloads) and that more referrals actually come from Facebook (0.66%, 5,126).  I was also pleasantly surprised to see a significant minority of traffic coming from Library Search (1.89%, 14,671).  But, really, ORO loves Google.

So where in the world do these external downloads come from… Audience location_v.2 Again, no surprises, the UK the biggest country of referrals (41.6%, 185,834), followed by the US (10.81%, 48,282) and India in third (3.42%, 15,278).

So, whilst the individual stories can be fascinating, the overall picture is that downloads come direct from Google and are most likely to originate from the UK.  The take away is we need to ensure ORO remains visible to search engines, it doesn’t actually matter whether people “use” the website, as long as they can access the content.