As it’s that time of the year and Altmetric has released its 2018 Top 100, in this post I share the 2018 Top 100 research outputs with ‘comics’ as a keyword according to Altmetric.
I queried the data from the Altmetric Explorer, looking for all outputs with this keyword between 13/12/2017 and 13/12/2018. I then refined the data to concentrate only on the Top 100 outputs about comics.
I am obviously very pleased to see The Comics Grid included in the Top 100.
It is interesting to note the diversity of countries associated to the profiles (where the metadata was available) giving attention to the outputs. According to Altmetric, there were 4,588 tweets about research outputs with ‘comics’ as keyword between 13/12/17 and 13/12/18 by 2,866 unique tweeters in 98 different countries. The map looks like this:
For more information and context on Altmetric and using the Altmetric Explorer, see my 2016 post here. Many other posts about alternative metrics and the Altmetric Explorer can be found throghout my blog.
Priego, Ernesto (2018): Countries and Number of Profiles that Gave Attention to Research Outputs with ‘Comics’ Keyword between 13/12/17 and 13/12/18 according to Altmetric. figshare. Dataset. https://doi.org/10.6084/m9.figshare.7467455.v1
Any frequent readers of this blog will be aware I am interested in article level metrics. I am particularly interested in the work done by Altmetric. Last week they published their annual top 100 list. I wrote this post about it.
The Altmetric Explorer is a tool for measuring the attention that scholarly articles receive online, and its intuitive user interface works as a live searchable database that allows users to browse the journals and repositories Altmetric tracks and obtain detailed reports.
On a weekly basis Altmetric captures hundreds of thousands of tweets, blog posts, news stories, Facebook walls and other content that mentions scholarly articles on the Web. The Explorer can browse, search and filter this data. The data can be exported by the user as ‘reports’ as simple text or spreadsheets, which can be then analysed in different forms. For example, The Explorer provides demographic data of the Twitter users found mentioning specific outputs, and thus works as a mechanism for the study of academic users of social media.
In the past few years I have often suggested, online, in talks, workshops and lectures, that the Altmetric Explorer can be useful to researchers as well. Librarians with access to the tool can help students and researchers get new views of recent articles that are receiving attention online. People often focus on ‘altmetrics’ as indicators of online activity around published outputs, but I often insist the Altmetric Explorer is useful as well as a tool for searching, discovering, collecting, creating, archiving, sharing and analysing bibliographic reference collections as datasets including not just bibliographic data including identifiers and/or URLs but also historical data of any metrics the service has tracked and quantified at the time of the data query/collection.
Inspired by Altmetric’s annual Top 100 list I used the Altmetric Explorer to search for the top articles with keyword ‘comics’ mentioned in the past 1 year. I did this particular search on the morning of Tuesday 20 December 2016. Dating the collection (and indicating the specific query) is always important as social media metrics are hopefully dynamic and not static (i.e. we expect an output’s altmetrics to change over time).
After my query I saved as usual my search as a ‘workspace’ on the app and then exported the dataset as a CSV file. I then manually cleaned and refined the data to obtain a file listing the top 100 references specifically on comics including their altmetrics. Data refining was needed to ensure the list included articles about comics, eliminating any non-relevant outputs (i.e. they were not about comics) and to correct text rendering errors, add missing data (like output titles when missing from the initial export) and limit the set to only 100 items by deleting the extra outputs.*
Hopefully it will be of interest to some of you out there. For comparison here’s these other datasets I have deposited on figshare in previous years:
Priego, Ernesto (2015): Almetrics of articles from the comics journals mentioned at least once in the past 1 year as tracked by Altmetric (20 August 2015). figshare. https://dx.doi.org/10.6084/m9.figshare.1514985.v3 Retrieved: 17 21, Dec 21, 2016 (GMT)
Though the two datasets above are outputs from different search queries (focusing on specific comics journals tracked by Altmetric rather than in any articles with keyword ‘comics’) we should we able to continue collecting data for future transversal studies.
Having yearly datasets obtained from the same queries, over a series of years, would provide evidence of comics scholarship’s presence online, and of the field’s (and Altmetric’s) evolving practices.
*It is possible the degree of relevance varies. Some outputs do not have ‘comics’ in their title but do discuss comics, for example ‘A randomized study of multimedia informational aids for research on medical practices: Implications for informed consent’ (Kraft et al 2016). It is possible however that a non-comics article or two remained, if you spot one do please let me know or leave a comment on the figshare output and I will correct and create a new version. It might also be noted that various outputs included are from The Conversation, which is not an academic journal, but it is tracked by Altmetric as it focuses on academic research news written by academics. For information and context about how Altmetric sources the data please read this.
This time the source data provides greater insights, particularly the article’s access type (Open Access, ‘Free’ or paywalled), type of content (article, letter, etc.) and subject.
Altmetric has already provided an analysis of this data (percentage of OA outputs in the list; countries of affiliations, institutions etc.) but having access to the source data means their analysis, visualisations and findings are actually reproducible (reproducibility was identified as a topic gaining interest; see Cat Williams’ post here). By providing access to the source data openly, other types of analysis are not only possible but encouraged (for example text and content analysis of the top 100 output titles).
One insight for me is that this list again demonstrates the dominance of the usual countries of affiliation, and up to a certain extent of the same journals (considering that Altmetric tracks a selection of publications, not all publications that exist).
I was interested in finding out whether the Top 100 would include any articles authored or coauthored by researchers with a Mexican institution as affiliation. There are two:
A genome-wide association scan in admixed Latin Americans identifies loci influencing facial and scalp hair features. Nature Communications7, Article number: 10815 (2016) doi:10.1038/ncomms10815 (Published online:01 March 2016)
Beverage purchases from stores in Mexico under the excise tax on sugar sweetened beverages: observational study. BMJ 2016;352:h6704 doi:http://dx.doi.org/10.1136/bmj.h6704 (Published 06 January 2016)
It is notable that both articles are the result of international coauthorship; the Nature Communications article including authors from other Latin American countries (Argenitna, Chile, Colombia); the BMJ one from Mexico and the United States. Importantly, both articles are open access.
I was also interested in seeing whether any Information Science or Computer Science research had made it into the list. There is only one article whose subject was categorised as “Information and Computer Sciences”:
Mastering the game of Go with deep neural networks and tree search. Nature 529,484–489 (28 January 2016) doi:10.1038/nature16961
This is a paywalled article authored by a team of 21 authors with Google DeepMind (London, UK) as affiliation.
I believe access to this data is useful to understand the evolving landscape of scholarly communications. It can also help us authors to gain insights into what kind of research is receiving attention online.
For example, the data seems to contribute to a body of encdotal and bibliometric evidence indicating that, for researchers with affiliations in ‘developing’ nations, open access and international collaboration remains key to greater visibility.
This year’s data also shows, again, that some countries (in the case of Africa, a whole continent), fields, and journals, remain under-represented or not present at all. It should also be noted that the only Computer Science article in the list is not by researchers affiliated to universities but to Google.
Yesterday I tweeted some quick thoughts after checking out the datasets, and compiled them using the new-ish ‘Moments’ feature on Twitter, which, for what it’s worth, I have embedded below.
I have done revisions to this post since publication.
[I don’t have time. What is this about?
My view is that altmetrics are not merely tools for the measurement of online attention but tools that can help us discover the literature that is being tracked as mentioned. I used the Altmetric Explorer as a tool to discover articles about inequality. I cleaned the data into three tables to reflect only the articles that interested me from three journals and then checked them for access and license type. Most are paywalled and if free access the licensing is not clear. Scroll down to see the tables, or download the dataset here.
It’s better if you read the post, though. ;-) ]
Using the Altmetric Explorer to Discover Literature
I‘ve been doing some research on the concept of ‘inequality’ from an economic and sociological perspective to add background to ongoing research on academic publishing and ‘monopolies of knowledge‘. I am interested in finding out more about the potential relationships between inequality of access to information (particularly access to peer-reviewed research publications) and other forms of inequality affecting social and economic development.
As you may (or not) know I am also interested in the potential for altmetrics as tools to help us in the discovery of research outputs. Some may not like it but needless to say people do search for and discover all sorts of information online. To give an example, these days many of us rarely get invited to a party with a paper invitation sent on the post (unless it’s a wedding, and even that is culture and country-dependent now); it’s likely, however, that there will be a Facebook invite, an Instagram account, or an email. OK, you may hate weddings or have never been invited to one. You must like music. If you are reading this you are likely to know people who discover new (and old!) music by looking into what other people listen to on apps like Spotify or Soundcloud, etc. (Yes, this sounds so old and so obvious!). We trust other people to recommend us stuff. (Think of how many of us travel today: TripAdvisor is a good example too).
More to the point, libraries and library web sites are no longer the only gateways to academic information (why should they be?). You don’t have to be a declared open education advocate to share, search for and discover interesting materials on Slideshare or YouTube. The distinction between ‘social networking’ or ‘social media’ sites and the rest of the Web is at best artificial: most platforms today imply inter-linking and therefore social interaction. Surely, I think, web platforms tracking social media activity like Altmetric can be used to discover what research people are mentioning online. One does not need a personal or institutional Altmetric account to discover other outputs from the articles themselves when they have Altmetric widgets. In other words, my view is that altmetrics are not merely tools for the measurement of online attention but tools that can help us discover the literature that is being tracked as mentioned.
The bibliography collection is an important part of a literature review. We may collect bibliography we are interested in reading before we properly review or collect as we read/review (hopefully once one is reading one follows leads in an article, checks the references and notes, clicks on links, gets elsewhere). To discover published research I have used the Altmetric Explorer many times before (see, as an example, “Ebola: Access and Licenses of 497 Papers Crowdsourced in 7 Days”, 14/08/2014).
Three Sets of Articles on Inequality
Recently I have been using it to search for articles on the topic of ‘inequality’. I am interested in which articles on this topic are being tracked by Altmetric as mentioned online, but I am also interested in the access and license types of the outputs tracked.
As I do normally in my research workflow I have been exporting the results of my searches and then cleaning the data. I do this by manually applying spreadsheet filters and adding and deleting columns, and using OpenRefine to deduplicate and standarise the data. I then check each output (i.e. I click on each link) and make a note whether I can access the full version without academic library credentials or not.
In this case I am sharing with you three sets of articles, each corresponding to a different journal that has published articles on inequality that have been tracked as mentioned online by Altmetric within the last year. In the tables below I have left the Altmetric score in timeframe (one year) in the first column and have organised the outputs in that order (from the highest score to the lowest). Having checked each article one by one manually not using any institutional credentials or IP, I have indicated in the last column the access type of each article. As Altmetric scores can change over time often quite quickly I have also left the most recent mention online according to Altmetric. This is of course not live data so it merely reflects the score and the most recent mention at the time of my data collection.
Information, Communication & Society
Altmetric Score in timeframe
Most recent mention online according to Altimetric
Racial formation, inequality and the political economy of web traffic
I am not sure if this humble blog would be tracked by Altmetric so (ironically) I may or may not be contributing to the Altmetric score of the outputs above as I am linking to them. (It is insightful that altmetrics can be tracked when people have reached merely abstracts but not full texts). In this instance I am not listing them above because I necessarily recommend them but as a small sample of articles on inequality from recognised journals, noting their access type.
I do not know if the authors of these articles have deposited open access versions of these papers in their respective institutional repositories or elsewhere (if you are so inclined, you can check the three journals’ archiving policies here), and I am not publishing this post because I cannot personally access the articles above (so thank you very much indeed but please do not contact me, dear reader, to offer me the PDFs via email or Twitter). I am not saying the articles above are all there is on the subject; I am just sharing those results and detailing their access type (which you can’t easily get unless you click on them and try to access them, and even if you can access them -this means full versions- you may find it difficult to tell why you happen to have access to them).
In this post I have wanted to make a very simple point: following the links to the publishers’ versions of record of these articles discovered via the Altmetric Explorer, the access conditions were the ones detailed above.
It could be argued that as an academic I have used the wrong tool to access these resources. It can be said that in my case, as an academic based in London, UK, it is my fault to expect to access these resources from outside my library (you say you can’t access them, dear reader? Your fault!) What I am trying to do here is try to see and share what happens when someone who normally has access to this kind of research steps out from their traditional/standard discovery tools and/or position of privilege. If you don’t have the right credentials, how much can you access? [I must also note that the Altmetric Explorer requires registration and normally membership too; however, all the links listed above can be reached via regular search engines and Google Scholar].
Things are changing slowly but academics’ distrust and complaints about the low quality and lack of trustworthiness of information found on the Web are common, but at the same time we have allowed paywalled online academic journals to remain (to me weirdly) disconnected from the rest of the Web, with links leading to abstracts that promise you a full version if you pay or have the right library credentials. This breaks the flow of information that has made the Web the amazing invention it is, and contributes to the separation between the outputs of higher education and the ‘general’ public.
In my opinion it is a serious problem that if you don’t have the right credentials then so much detective work is required to access some important research (or to elucidate articles’ licensing conditions, even if they are ‘free’ or ‘complimentary’). Others, as we know, can’t be bothered at all and merely jump all the hoops, against all policies. The more barriers you impose, the more people will want to circumvent them. Ideally.
In reality, it is more likely that paywalled outputs remain inaccessible/invisible to the larger public, and perhaps even more to those affected by the very conditions studied in them. Even as an academic or student in an elite institution it is often hard (read: not straight-forward, not friction-free) to access them! A non-academic searching for this research online is likely to have already transcended many of the structural barriers created by inequality. Once you finally get to an interesting article, how great it must be then to be greeted by a huge ‘pay or keep off’?
Some might say my hypothetical non-academic individual seeking access does not really exist. Some have suggested to me that there is no evidence there is interest from the public, and that those who have access are the only ones interested. That the non-academic public wouldn’t understand the research anyway. That those interested could try harder to find surrogates. That in case they exist they are likely to know people who can ‘share’ the research with them anyway. The list of justifications of the current system can be long.
Having lived, studied and worked in a developing country I know intelligent, curious, well-informed bilingual individuals who have no access to versions of record do exist. This is people who face the inequalities of access to scientific information. They may be relatively privileged, because they have transcended the most pressing needs to enable them to seek out research. This, however, does not mean they do not exist and that their needs are not important.
I know interested individuals that are not academics exist here in the UK too. I also know for a fact that there are academics worldwide who do not have access to a lot of paywalled research. I am often one of them myself. I know there are others because I know them personally and because we know that not all libraries can afford to subscribe to the same ‘bundles’ (for the latter there is a growing body of evidence). My personal experience does not count as scientific evidence, but it matters to me and I know it matters to others. I question why we assume that if there is supposedly no current public demand for research then it is acceptable to paywall it and not encourage further public interest and demand.
I am aware it is getting boring because I have been repeating this for several years know, but legal ‘frictionless sharing‘ wouldn’t go amiss, especially for this type of research. We call it “open access”.
Priego, Ernesto (2016): Inequality: Three sets of Journal Article Titles and URLs/DOIs from Three Different Journals, with Altmetric Score in Timeframe (1year), Last Mention at the Time of Collection and Access Type Noted. figshare. https://dx.doi.org/10.6084/m9.figshare.3808134.v2 [CC-0].
Euan Adie from Altmetric has published a very interesting article with insights into a dataset of Nature Communications articles published between October 2013 and October 2014. I uploaded an edited version of his set as a spreadsheet on figshare.
Open access articles, at least those in Nature Communications, do seem to generate significantly more tweets – including tweets from people who tweet research semi-regularly – and attract more Mendeley readers than articles that are reader pays.
I took his dataset from figshare and opened the .txt file using Excel. Using filters I deleted the 2013 articles and only focused on the ones published between January and October 2014. I sorted them by month of publication from January to October and separated them into two sheets, one for the open access articles and another one for the paywalled ones. This way one can use this spreadsheet to access the open access articles without having to sort through the paywalled ones. Or one can just do an individual analysis per type of access. This is of course super rudimentary data refining (if it can be called that), but it helped me to focus on the differences between access types published this year.
The resulting edited file has two sheets organized by type of access and month of publication, including only articles published during 2014. (Note: seven (7) records appearing incorrectly as published 1900-1 under the month column were removed. They had no online mentions).
Manual manipulation of the original data was performed so all data should be contrasted with the original source cited above.
The intention of sharing this edited file is to aid in focusing on the open access and paywalled outputs published between January to October 2014 as provided in the original dataset. Having a smaller dataset organized by date and type of access may make quick visualisations easier.
I have shared a file containing a sheet with a list of 8, 348 journal articles obtained from a basic search for the keywords “racial” and “etchnicity” in research papers mentioned online anytime as tracked by Altmetric.
The file contains a sheet with the list of all 8,438 bibliographic entries, including DOIs and URLs, and mention counts on different Web services as tracked by Altmetric.
The entries are ordered from the highest-scoring (best quality of online attention) down. The Altmetric score is a quantative measure of the quality and quantity of attention that a scholarly article has received.
I have edited the spreadsheet adding columns H, I and J, hoping any users of this data are interested in adding the type of access, license and, if applicable, price of each output.
This file has been shared with the intention of creating awareness of the scientific/academic literature mentioning the above-mentioned keywords being mentioned online as currently tracked by the altmetrics service employed to obtain the dataset. Data might require refining and deduplication.
Please note that the academic disciplines and methodologies represented in this dataset reflect the sources curated by the service employed. This is an unedited report obtained through a basic automated search so not all entries might be considered relevant and users will require to refine the data to fit their own needs. There is some very interesting and useful stuff there.
I created and shared this file with a Creative Commons- Attribution license (CC-BY) for non-profit academic research and educational use.
¿Qué hice hoy? Entre otras cosas (entre las cuales siempre se cuenta contestar unos 50 mensajes de correo electrónico al menos, todos de trabajo) me propuse intentar hacer una “infográfica” de un reporte que hice con el Altmetric Explorer sobre artículos académicos que mencionan la palabra “mastectomía” en el título.
Cuando se anunció en la prensa que Angelina Jolie había optado por la mastectomía doble como medida preventiva, hice un reporte inicial esperando poder comparar en el futuro si el interés público por el tema causaba o la aparición de nuevos artículos académicos o nuevas menciones en línea de artículos académicos previamente publicados sobre la mastectomía. Ese reporte inicial quedó congelado en el tiempo en un archivo fijo que subí a Figshare, aquí:
Hoy con el nuevo reporte que hice noté que dicho y hecho un nuevo artículo con “Angelina Jolie” en el título había hecho su aparición no sólo entre los diez más mencionados sino entre los tres más mencionados.
La itención fue ver qué tan rápido podía armar una infográfica para compartir en línea (no para imprimirse, es decir en baja resolución y hecha con software gratuito en línea) con estos nuevos datos. Me tomó buena parte de la mañana y mi computadora falló varias veces en el proceso. Al final no quedó como me hubiera gustado pues se fueron algunos errores menores de redacción y de puntuación, pero bueno, se comparte abiertamente de buena fe y quien quiera enmdendar o hacer mejor las cosas para eso se comparte con licencia de CC- Atribución.
I created this infographic which presents some findings from an Altmetric Explorer report I retrieved on 29 May 2013 9:59AM.
Needless to say I am not a professional designer and don’t have professional image-making software. Nevertheless I gave it a go to see what I could come up with quickly. I thought it had to be done and shared rapidly as it represents a snapshot in time (social media mentions can change over time).
It is a low-res .png made to be seen and shared online.
I spotted some typos and a sentence that would require rewriting but it took me so long to produce this version I will have to let it be. The CC-BY license means anyone could –in theory– correct it if they so wished.
My new post at Altmetric is guided by the concept of digital opportunity. I take a quick look at some research on the uptake of ICTs and social media in Africa and link to and comment on Altmetric details of two articles on Africa we located through the Altmetric Explorer.
As a Latin American I have always been wary of the risk of mis-representation in research on developing nations which is carried out from a developed-nation perspective. On the one hand I believe we always function from specific positions that enable us or disable us, empower us or disempower us; on the other hand I believe it is possible to at least try to recognise all that we can’t possibly know given those positions. In other words this is to say that this is a post that was very difficult to write for me. I attempted to back up any claims with the consulted research and to offer a balanced, yet accessible perspective. It is a blog post, not a research paper, so what I can and cannot do is also determined by it.
I would like to thank Euan Adie at Altmetric for his patient critical feedback, Tomi Oladepo, (Researcher on Digital Public Sphere, University of Warwick) for her kind help locating some bibliography and to Julie Soleil Archambault (Departmental Lecturer in African Anthropology, Oxford University) for answering my questions.