All 3 entries tagged Statistics
View all 61 entries tagged Statistics on Warwick Blogs | View entries tagged Statistics at Technorati | There are no images tagged Statistics on this blog
February 10, 2020
Writing about web page https://exchanges.warwick.ac.uk/issue/view/27
Today I’ve been having a quick exploration of the numbers around the most recent couple of issues of Exchanges. Metrics are for (most) journals a hotly contested topic, with their value quantitatively established and promoted with profound pride by editors and publishers alike. Regular readers of this blog will be aware than I’m from the qualitative school of research and have some deep ideological objections to the metrification and quantisation of academic publication, and consequential transformation in a highly fetishised quasi-market mode. Ahem. To read more on this topic, see my publications or come and have a chat with me, although I cannot promise not to get onto my soapbox!
Personally, I’d rather see the valorisation of an article through post-publication discourse in the social and public spheres, than watch the uptick of citations or downloads. However, for most of our authors and readers alike metrics and journal publications, love ‘em or loathe ‘em, are intrinsically linked. I can appreciate being able to see how people are reading the work author’s have slaved over for months, in an employment sector often detached from tangible esteem measures, can be a key personal satisfier.
As an editor-in-chief too, I confess I do get a little frisson of delight watching the download statistics slowly (and not so slowly) grow post-publication . For the authors, seeing these figures climb mean people are at the very least reading their publications, although how they are using it, citing it, teaching from it, learning from it, remain to be elicited. As a journal publisher, it helps me to promote the journal as a publication destination for future authors, and to answer questions to my employers about the continued viability of the title.
In recent issues of Exchanges, we’ve shifted to include more ways within articles to recognise and identify authors, notably ORCIDs, twitter handles and biographical sketches. This means it has become easier to spot a portion of the buzz around an issue and its concomitant articles. Certainly, Vol 7.2 (Cannibalism Special Issue) has generated a highly visible amount of discussion following its publication, which I hope will continue as more people read the issue . Including author twitter handles means I’m at least able to observe part of these conversations, even though monitoring discussions within departments, conferences and the like isn’t practical. I fervently hope this most exciting issue will continue to receive a suitably wide discussion, as we continue our promotional efforts over the next few weeks .
But back to my original point: metrics. I was curious this morning, now we’re just over 10 days post publication, to see how the issue was progressing. So, I ran some quick, back-of-the-envelope calculations to contrast v7.2’s readership numbers with the preceding issue. Here’s what I found.
Table 1: Download stats/article for the two most recent issues of Exchanges 
|Issue||V7.2 (Cannibalism)||V7.1 (Regular issue)|
This is by no means conclusive but these numbers suggest the level of interest in this issue is potentially above the norm for Exchanges. If this degree of reader engagement continues, it wouldn’t surprise me if after 100 days post-publication most of this issue’s articles will have developed an especially commendable download rate. Kudos to the authors, who through being associated with such a broad, critical mass of learned discourse, will be able to reap additional benefits. I’ll certainly endeavour to return down the line to see if my assumptions are being met.
Additionally, this also suggests how adopting a publication mode which embraces more special issues such as this one can be considerably beneficial to Exchanges health and longevity as well. The more readers we garner, the more likely people will cite the articles, helping enhance the title’s valorisation and recognition, which in turn encourages more submissions. As the managing EIC, right now, I couldn’t be more delighted with how all these efforts have turned out. Even if it has substantially increased my own workload!
 Incidentally, my own IP is masked from the stats, so it doesn’t matter how many times I open or download an article to check something, my interactions aren’t actually counted. But then, I’ve already read each article a handful of times already on its journey to publication…
 I’m still keen to develop post-publication commenting functions for readers and authors on our journal site, but currently, am awaiting an update to the platform before I can make any strides in this direction. If you’re one of our readers, authors or reviewers and you’d like to see article comments; drop me a line – as the more people who ask, the more I can lobby my lovely technical team to devote some time to it!
 Another medium-to-long-term goal is to introduce altmetrics scores for each article, to try and capture a value for how ‘talked about’ each issue is in the public domain. Watch this space for details as soon as I have them, but I can assure you, this is one of my ‘top 5’ goals for Exchanges in 2020.
 V7.2: 11 days post-publication, V7.1, 101 days post-publication
August 21, 2018
Over the weekend I had a brief discussion with the other Dr Johnson in my house, who was asking me various questions about Exchanges, its metrics and readership. Okay, truth be told I started the conversation by wondering aloud about various aspects of our multiple audiences , as it’s a topic never too far from my mind, even on a road trip to the far south of the UK. I should mention, Mrs the Dr Johnson is a remote-sensing satellite and environmental monitoring specialist at another Midlands university, and I suspect tends to perceive the world through a geographic lens. Hence the construction of her question and interest. I’ll confess it wasn’t something I could immediately answer while driving down the M40, beyond making an assumption that our to-date core audience was located in and around Warwick, and perhaps Monash, given our concentration of editors and authors from those locations.
As I’ve discussed before, one of my (many) ambitions for Exchanges is to broaden the range of its audiences , and thinking about what we can find about the current audiences isn’t a bad place to start. I’ve two principal tools at my disposal for gathering this sort of data: Google Analytics (GA) and the Open Journal System’s (OJS) inbuilt statistics generator. The former looks a lot slicker and can churn out some pretty illuminating graphics at the click of a mouse, the latter’s UI and outputs are a lot more ‘web 1.0’ - in that creating a custom report is not a facile exercise and the platform spits out reams of largely unformatted, hard, numerical data. Both tools have their places in my working practices, for example at times it’s handy to have access and manipulate raw data, and GA doesn’t make scraping that in its entirety quite as easy. Conversely, when I need an illustrative graphic in short order for a presentation or report, GA is the tool I turn to.
The $64,000 question: does their data correlate? The answer is yes…and no. Broadly there’s some alignment, but the figures each one has presented me with are reasonably different in exact value if similar in relative magnitude. Given the issues with generating comparable data over the same period  it comes as no surprise to me that variance in ranking beyond the ‘big three’ UK, USA and Australia  exists. Perhaps more interesting are those countries which appear in one but not the other analytical tool’s top 10.
|Google Analytics||Open Journal Systems|
|1||United Kingdom||United States|
|2||United States||United Kingdom|
|8||South Korea||South Korea|
(countries appearing in both lists highlighted)
This might suggest, given GA has been running for less time than OJS’ current platform, that Canada and the Philippines are new and expanding audiences for Exchanges, with France and Russia diminishing. However, the precision in the time spans over which this data was gathered are both too limited to make such sweeping conclusions . It is pleasing to see some non-Anglophone usage though in both charts, especially considering our sole language of publication is English.
Anyway, no matter the deeper implications of this very light touch look at Exchanges’ user statistics, I think I’ve at least answered part of Mrs Dr Johnson’s question about from where in the world our usage has originated. Naturally, this beggars another question which I can’t immediately answer: where SHOULD our audiences for the journal be coming from? As always, answers in the comments below please…
 Readers, authors, potential authors, stakeholders and more…I’m still work on defining these
 Does the ISS have an ISP I can track? If it has, another mission is to get this journal read in orbit!
 These issues are multiple. For examples, with OJS, when we moved to the newer version last year this, regrettably, seemed to ‘reset’ the statistics for the platform. We’ve a back record of these, but it’s no longer possible run off a complete set since the journal began. Likewise with Google Analytics, we’ve not had this running the whole time the platform has been up, so there’s going to be a temporal discrepancy there too. Added to that neither platform counts or creates its statistics in the same way, without a LOT of lengthy post-processing and normalisation, for normal usage there are always going to be disagreements on the ‘exact’ magnitude of visitations. Just one of the reasons as a qualitative researcher, I tend to maintain a certain analytical cynicism wherever ‘statistics’ are used to justify something: there’s always likely flaws, assumptions and simplifications in the underlying data acquisition methods!
 These make up 58.4% (GA) or 77.7% (OJS) of all usage
 It is possible I could make the data collection time frames marry better, but I’m still developing an understanding on how OJS works ‘under the hood’ in this respect. Something to return to at a later date, perhaps.
August 07, 2018
The month of August is the time of year when, traditionally, UK universities slow down a bit. It’s the summer holidays, so staff with families take the time to go on vacation, meaning many an email goes unanswered for a while and progress can seem sluggish. Personally, as a former academic librarian, August was the month I was often the busiest as all those project tasks and new academic year preparatory efforts were always in full swing! As Senior Editor of Exchanges, I felt this slight pause in email traffic and article ingest was the perfect time to push ahead on some the developmental work and background research advancing the journal for the benefit of our authors and readers alike.
The most exciting new addition is that you can now see the downloads from the past year of all the articles on Exchanges. This is a great way for authors to track their papers’ engagement, not to mention for the Editorial Board to identify the areas in which our readers are most interested. This is our first step towards providing more information about how well the material published with us is being received, and over the next year, hopefully, I’ll be able to highlight further new information in this area.
Secondly, coming very soon will be DOIs for every article on Exchanges, both past and future. DOIs (digital object identifiers) are a unique alphanumeric string which provide a persistent lifetime link to a particular location on the internet, a shorthand if you like, for each article on Exchanges. This means even if we alter the journal’s website location or change our domain address, the DOI will remain a stable and viable way for readers to access an article. Additionally, I think they also make citing articles look a little tidier.
From today, we’ve also changed the Creative Commons author publishing licence for Exchanges, from the more restrictive Attribution-Non-Commercial-Sharealike (CC-BY-NC-SA), to the more desirable Attribution (CC-BY) only. The prior licence was considered the bare minimum to meet funders and governmental agencies around the world’s open access requirements. Shifting to a CC-BY licence brings us more in line with the major interdisciplinary titles, such as PLOS One, and further demonstrates Exchanges’ adherence to no-author-fee diamond model  open access publishing. Previously published, or submitted articles, will retain their original licences, as agreed by their authors. Newly submitting authors from now on, will be asked to accept the new licence terms at the point of submission, as part of their publishing agreement with Exchanges.
Behind the scenes I’ve also rolled out our very first author feedback system. This ties into my previously discussed interest in our author and reader audiences, and will provide some initial data towards satisfying that curiosity. It will also contribute by identifying aspects of Exchanges’ platform and process which work well, or less so, for our authors, directing my attention to where the greatest benefit can be achieved. If you’re one of our prior authors (vol 5.2), and you’ve had one of my emails about this, please do respond as it’s a very short set of questions which won’t take a lot of time. If you’ve already responded, many, many thanks!
Finally, a big welcome to the new members of our Editorial Board, who I know will make a considerable contribution to the running and engagement of the journal, and I’m really looking forward to working alongside them. Hopefully, I’ll be able to announce a few more additions to the team in the coming months, as the title continues to grow.
Hence, as you can see, the summer is continuing to be anything except quiet for Exchanges!
 As per Fuchs and Sandoval, although some might call this the ‘radical mode’, if they’re more a fan of Gary Hall’s work.