Does presence of social media plugins in a journal website result in higher social media attention of its research publications
TThis is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/
Does presence of social media plugins in a journal website result in higher social media attention of its research publications?
Mousumi Karmakar a , Sumit Kumar Banshal b & Vivek Kumar Singh a,1 a Department of Computer Science, Banaras Hindu University, Varanasi-221005, India b Department of Computer Science, South Asian University, New Delhi-110021, India
Abstract:
Social media platforms have now emerged as an important medium for wider dissemination of research articles; with authors, readers and publishers creating different kinds of social media activity about the article. Some research studies have even shown that articles that get more social media attention may get higher visibility and citations. These factors are now persuading journal publishers to integrate social media plugins in their webpages to facilitate sharing and dissemination of articles in social media platforms. Many past studies have analyzed several factors (like journal impact factor, open access, collaboration etc.) that may impact social media attention of scholarly articles. However, there are no studies to analyze whether the presence of social media plugin in a journal could result in higher social media attention of articles published in the journal. This paper aims to bridge this gap in knowledge by analyzing a sufficiently large-sized sample of 99,749 articles from 100 different journals. Results obtained show that journals that have social media plugins integrated in their webpages get significantly higher social media mentions and shares for their articles as compared to journals that do not provide such plugins. Authors and readers visiting journal webpages appear to be a major contributor to social media activity around articles published in such journals. The results suggest that publishing houses should actively provide social media plugin integration in their journal webpages to increase social media visibility (altmetric impact) of their articles.
Keywords:
Altmetrics, Science Communication, Social Media Attention, Social Media Plugin.
Introduction
Scholarly articles are now disseminated and shared on different social media platforms such as Twitter, Facebook, LinkedIn as well as academic social networks like Academia, Mendeley and ResearchGate. These social and academic social networks provide wide-range of facilities which can be useful for academics (Gruzd & Goertzen, 2013), with some studies (Williams & Woodacre, 2016) pointing out that these social and academic networks are informative and relevant for quantitative characterization of research assessments. The activity of measuring the interaction of social media with scholarly information processing is now a well-established area, known as altmetrics (Priem, 2014; Priem et al., 2010; Priem & Hemminger, 2010). Over the last few years, altmetrics has attracted a lot of attention from different quarters, including research performance assessors, authors and journal publishers. Owing to the popularity of altmetrics and increased social media attention to scholarly articles, many journals have now integrated social media plugins in their web pages to facilitate sharing and dissemination of articles in social media platforms. Corresponding Author. Email: [email protected] his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/
Several previous studies on altmetrics have tried to analyze the correlation between altmetrics and citations, many providing the evidence that altmetrics are positively correlated to citations. These studies used data from a variety of social platforms and/ or academic social networks and found a varying degree of correlations between altmetrics and citations, ranging from weak positive to strong positive. Data samples drawn from different countries/ regions, journals and subjects were analyzed in different studies and the overall understanding is that there exist positive correlations between altmetrics and citations. Due to the increasing social media activity around scholarly articles and the positive evidence of the correlation between altmetrics and citations, many studies have also tried to analyze what factors or article characteristics may result in higher social media attention of scholarly articles. The factors analyzed in previous studies include journal impact factor, international collaboration, open access etc. Studies have suggested that international collaboration and open access are both found to be positively associated with higher altmetric impact of articles, whereas no conclusive evidence could be found about journal impact factor. One important factor that has, however, remained totally unexplored is the presence or absence of social media plugin in journal webpages. Journals are now integrating different social media plugins directly in their webpages to facilitate social sharing and dissemination of information about articles. Many journal publishing houses now provide for integration of well-known social platforms like Facebook, Mendeley, LinkedIn, Twitter etc. in their journal webpages. These plugins are usually integrated in journal websites in the form of icons that contain a link to the respective social media platform. By clicking the link, a user may easily tweet, like, share or comment on an article in the journal. As a practice, journals which have these plugins integrated also provide different kinds of counts and scores about social shares/ mentions of articles. It would therefore, be interesting to analyze what could be the impact of these plugins on social media visibility of articles. In the absence of any studies on this aspect, it is not clear whether the presence of a social plugin in journal webpages actually results in higher social media activity around articles in that journal. This paper aims to explore this question by analyzing a sufficiently large-sized sample of articles published in journals of both kinds: those with social media plugins and those without the plugins. The journals used for the analysis are selected in a manner that all major disciplines and publishing houses are represented. The coverage and average mention values in different social platforms for the articles in these two sets of journals are analyzed to observe the possible impact of the presence/ absence of social media plugins in journal webpages.
Research Question
The main research question being explored is:
RQ:
Does presence of a social media plugin in a journal result in higher social media attention to articles published in that journal?
Social media data for a large-sized data sample comprising of 99,749 articles drawn from 100 journals, 50 with social media plugins integrated and 50 without the plugins integrated, is analyzed to find an answer to this research question. his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/
Related Work
Over the last 6-7 years, different kinds of research studies have been carried out in altmetrics ranging from country-specific studies (Banshal et al., 2018, 2019a; Hammarfelt, 2014) to discipline-specific studies (Bar-ilan, 2014; Chen et al., 2015; Htoo & Na, 2017; Sotudeh et al., 2015; Vogl et al., 2018). Some previous studies also tried to analyze the connection between discipline and social media attention levels of scholarly articles (Banshal et al., 2019b; Costas et al., 2015b, 2015a; Ortega, 2015; Thelwall & Kousha, 2017) and found that disciplinary variations exist in social media coverage of scholarly articles in different platforms. One of the most researched questions in altmetrics research, however, has been to understand the relationship between altmetrics and citations. There are several studies conducted on data from various social media platforms as well as academic social networks, with the objective of understanding the relationship between altmetrics and citations. Among the initial studies is the work by Eysenbach (2011), in which Twitter data for publications in the Journal of Medical Internet Research were analyzed and it was found that tweets could be an early predictor of citation counts in the domain of medical sciences. A set of 286 research articles constituted the data for analysis and immediate tweets were analyzed. Shema, Bar-IIan, & Thelwall (2014) analyzed two different sized journal samples (12 & 19) for two different time periods (2009 & 2010, respectively) and found that blog citations correlated with citation counts for some journals. Another study in the same year by Haustein et al. (2014) analyzed 1.4 million biomedicine articles indexed in PubMed and Web of Science, published during 2010 and 2012 in more than 5000 bio-medical journals, and found that tweets have weak correlations with citations. They further found that less than 10% of articles from PubMed were found to have a tweet in the collected dataset. Several other studies over the time also worked on understanding the relationship between altmetrics and citations. Sotudeh, Mazarei, & Mirzabeigi (2015) worked with CiteULike bookmarks for Library & Information Science (LIS) papers and found that they correlate well with citations. Out of 83 source titles examined between 2004 and 2012, ~78% titles had at least one article bookmarked in CitedULike. The citations of LIS papers were found to be weakly but positively correlated with CitedULike bookmarks. Costas, Zahedi, & Wouters (2015a) analyzed a dataset collected for the period of 2005-2012 and confirmed the claims of positive correlations between altmetrics and citations, but found that these correlations are very weak. They also found that only 15-24% of articles get mentioned in social media platforms. Maggio et al.(2018) worked on data of the health professional education (HPE) domain and observed that altmetric indicators are not necessarily found to predict higher citation, but they can help to generate more citations and visibility. They evaluated around 2500 HPE articles published during 2013-2015 in the study. They have also shown that blogging was the most impactful alternative medium of propagating citations. Thelwall (2018) worked on Mendeley reads and found that Mendeley reads have a high correlation with citation counts for ten selected disciplines. In another work, Thelwall & Nevill (2018) examined 27 research areas classified by Scopus and analyzed the data for 10,000 selected articles. They found positive correlations between Mendeley reads and citation counts. Another interesting direction of research in altmetrics, which perhaps is more closely related to the broader theme of present work, has been to identify various factors or article characteristics that may be positively associated with higher altmetric attention of the research articles. Different factors like journal impact factor, authorship structure, length of article, international collaboration, open access etc. have been analyzed for this purpose. One of the first work (Haustein et al., 2015) on the theme was focused on studying the effect of document his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/ properties (mainly discipline, document type, title length, number of pages and references) on social media metrics of scholarly articles. They found that collaboration leads to higher impact. However, they observed that different article characteristics are not similarly associated with bibliometric as well as altmetric impact. For example, they found that the higher impact factor of a journal is positively related to the bibliometric impact of its papers, but it may not be similar for altmetric impact. They have concluded that factors driving social media and citations are different. Another study (Didegah et al., 2018) observed the differences in the role of selected factors on citations and altmetrics by analyzing a sample of Finnish articles. They found a positive relationship between journal impact factor and bibliometric & altmetric impact, particularly on some platforms. The collaboration was also found to have a positive role in both kinds of impacts. In another work in the same year, Zhang & Wang (2018) observed that in the case of biology, journals with higher impact factor values are not found tweeted often, i.e. correlations are not seen. Regarding studies on the association of open access (OA) with social media attention, some studies (Davis et al., 2008; Gargouri et al., 2010; Norris et al., 2008) studied its association with bibliometric impact, whereas several studies (Poplašen & Grgić, 2017; Wang et al., 2015) tried to look at its association with altmetric impact and found a decisive advantage for OA articles. In general, all the studies concluded that open access journals/ articles get higher bibliometric as well as altmetric impact. One recent study (Holmberg et al., 2020) used data for Finnish universities and showed that there are significant disciplinary and platform-specific differences in the “OA advantage”. They found that articles in OA journals in disciplines like veterinary sciences, social & economic geography and psychology receive more citations and attention on social media platforms, while the opposite was found for articles in OA journals within medicine and health sciences. The previous studies on factors and article characteristics, thus, could help in understanding the impact that these factors or characteristics may have on the altmetric attention of research articles. It was observed that international collaboration and open access are positively associated with higher social media attention of scholarly articles. The journal impact factor, however, was not found to be positively associated with higher social media attention in all cases. The authorship structure’s association with social media attention of articles, however, more or less remains unexplored. Most studies on authorship structure essentially targeted international collaboration of authors and the effect of the number of authors on altmetric impact of articles is not very well studied. Despite existence of several studies on the impact of different factors/ article characteristics on altmetric impact of articles; the association of presence/ absence of social media plugins in journal webpages on the altmetric impact of their articles, has remained totally unexplored. As a result, it is not clear whether the presence of social media plugins in journal webpages could have a positive impact on social media attention of research articles in those journals. To the best of our knowledge, there is no previous study to have studied this question. As a practical consequence, it is also not known whether integration of social media plugin in journal webpage can have a positive impact on altmetric attention or not. This article aims to bridge this gap in knowledge by analyzing a sufficiently large sized data sample. his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/
Data
The data for analysis was obtained from two sets of journals, those with social media plugin integrated in their webpage and those that do not have such integration. Two data sources have been used to obtain publication and social media data. The research publication data of two sets of 50 journals each, as described ahead, for the year 2016 was obtained from the Web of Science database. The social media data for each research publication was obtained from Altmetric.com for three social platforms- Twitter, Facebook and Mendeley. The first set of journals comprised of 50journalsthat have social media plugins integrated in their webpages, for at least two social platforms (out of three platforms- Twitter, Facebook &Mendeley- analyzed by us). The second set comprised of 50journals that do not have plugins integrated for any of the three social platforms. We looked at about 300 journals from different publishers in the process, finally selecting only those that met the criteria of presence or absence of social media plugins in their webpages. In the process of selecting journals, we tried to have a suitable representation of different major publishers as well as disciplines. These two sets are, hereafter, referred to as SMP and NSMP, where SMP refers to set of journals having social media plugin, and NSMP to journals that do not have social media plugins. The SMP set comprised of journals from nine different publishers, namely Elsevier (9 journals), Wiley (9 journals), Oxford Academic (8 journals), Taylor and Francis (7 journals), Nature (5 journals), CellPress (4 journals), PLoS (4 journals), SAGE (2 journals) and ACS publications (2 journals). The NSMP set comprised of journals mainly from four publishers- Springer (16 journals), IEEE (19 journals), and Hindawi (8 journals), and Emerald (7 journals). These journals taken together represent all major disciplines of research. This was an important thing for us to ensure, as it has been seen in earlier studies that articles in some disciplines get slightly higher social media attention than average. Therefore, having representation of all major disciplines will help in neutralizing any extraneous effect of discipline. After selecting the journals for both SMP and NSMP sets, the publication records for each journal for the year 2016 were downloaded from the Web of Science database. The 50journals in the SMP set, taken together, account for 78,633research publications and the 50journals in the NSMP set, taken together, account for 21,949research publications. All kinds of research publications (different document types in the Web of Science database) were included in the data for analysis. Since all the articles did not have a Digital Object Identifier (DOI), which was the essential linking factor with altmetric data, we had to drop publication records without DOI. Removing publication records without DOI left us with 77,820 articles (with DOI) in the SMP set and 21,929articles (with DOI) in the NSMP set. The analytical study, thus, used a sample of total 99,749 articles drawn from 100 different journals representing 13different publishers. Tables 1 and 2 present the detailed data about the journals used in the analysis. For each publication record, as obtained above, having a DOI, the social media data corresponding to three platforms- Twitter, Facebook and Mendeley- was obtained from popular altmetric aggregator- Altmetric.com. ADOI lookup was performed in Altmetric.com explorer to get the coverage and mention data for publication records for the three platforms. Out of 77,820 publication records in SMP set, 50,084 publication records had some mentions in Twitter; 13,457 had some mentions in Facebook; and 29,117 had some mentions in Mendeley. Similarly, out of 21,929 publication records in NSMP set, 4,677 publication records had some mentions in Twitter; 1,050 had some mentions in Facebook; and 7,791 had some mentions in Mendeley. his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/
Methodology
The social media coverage and mention data for research publications/ articles from the two sets of journals were analyzed computationally by writing Python programs. For each journal, the coverage percentage of its articles in all the three social media platforms is obtained by identifying the proportion of articles that get some mentions in respective platforms. The coverage values for all articles in a journal are averaged. These average values of coverage for all the journals in a set are then aggregated by computing the mean and median of coverage percentage values. This is then done for the NSMP set as well. In addition to coverage percentage, mentions per paper is also computed for each journal for all the three platforms. For example, in order to compute average mentions per paper in the Facebook platform for a journal, the total number of mentions for all the articles in the journal are divided by total number of articles in that journal. This way, the average mentions per paper is computed for all the journals in the two sets- SMP and NSMP. The journal-wise average mention values in each set are then aggregated into a single value summary by computing the mean and median values. This single value summary is taken as a representative value for each of the sets, SMP and NSMP. Thus, coverage and average mentions are the two main parameters computed to distinguish between SMP and NSMP sets and to observe whether SMP set gets higher coverage and average mentions as compared to NSMP set.
In order to understand whether the average mentions per paper are suitable to differentiate between SMP and NSMP journals or not, we employed area under curve (AUC) values as a measure. We plotted receiveroperating characteristic (ROC) curves to identify the association between individual mentionsof articles for each platform (Twitter, FB, Mendeley) and journal class (SMP or NSMP). Primarily, ROC-AUC procedure is used to test/ compare the performance of one or more classification algorithms, but we have utilized this mechanism to analyze the association strength of features and classes as this was found useful earlier (Hassan et al., 2017). The Logistic Regression model is used to classify the articles published in collected journals. For classification purposes, all publications in the SMP and NSMP sets are tagged with 1 and 0, respectively. The classification task is performed using the built-in Logistic Regression model of sklearn.linear_model module available in python. For each of the three platforms, publication records are classified separately based on their respective mention value to measure the discriminative capability of that platform. Two important parameters involved in the ROC-AUC mechanism are Sensitivity (True Positive Rate or TPR) and, Specificity (1-False Positive Rate or 1-FPR) . These terms are defined as follows 𝑆𝑒𝑛𝑠𝑖𝑡𝑖𝑣𝑖𝑡𝑦 = 𝑇𝑃𝑇𝑃 + 𝐹𝑁
𝑆𝑝𝑒𝑐𝑖𝑓𝑖𝑐𝑖𝑡𝑦 = 𝑇𝑁𝑇𝑁 + 𝐹𝑃 where, TP = True Positive (actual and predicted, both values are positive), FP = False Positive (actual value is negative but prediction value is positive) , TN = True Negative (actual and predicted, both values are negative), and FN = False Negative (actual value is positive but prediction value is negative) Various parameters of ROC curves are computed using two functions, present in the sklearn.metrics module of python. The values of true positive rate, false-positive rate and thresholds are computed using roc_curve function, whereas the AUC measure is found using the roc_auc_score function. We have https://towardsdatascience.com/understanding-auc-roc-curve-68b2303cc9c5. Accessed(10/01/2020) his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/ used plt function of a python library matplotlib.pyplot to plot the curve with the previously computed parameters of ROC. ROC curves are plotted for each of the three platforms separately, to identify their discrimination strength. Higher the association between mentions and journal class, higher is the distinction capability. An AUC, which is close to 1, shows good separability of the algorithm under consideration as well as the association between the discriminating feature and the target class. In order to find the extent of such relationship, a logistic regression model was deployed, as detailed above. Tweets, Facebook mentions, and Mendeley reads are taken as independent variables while the class label is the dependent variable. The AUC computations confirmed the discriminative nature of mentions for the given task and hence results are found to be useful for the question analyzed. Results
The social media attention levels of all the articles in the two data sets were analyzed to identify if there exist identifiable differences in the coverage and attention levels between articles published in journals having social media plugins (SMP set) and journals that do not have the plugins (NSMP set).
Differences in overall social media coverage of journals in SMP and NSMP sets
The first observation is with respect to coverage of articles in journals of the two sets. Tables 1 and 2 show the overall social media coverage levels of articles, as per the data obtained from Altmetric.com, for the SMP and NSMP sets, respectively. The coverage percentage here implies the proportion of articles from a journal that get some social media activity. It can be observed that in the SMP set, 35 out of 50 journals have more than 50% articles covered in the Altmetric.com, whereas in the NSMP set, only 14 out of 50 journals have more than 50% coverage. The mean coverage percentage for the SMP set is 71.556%, whereas the mean coverage percentage for NSMP set is 36.677%. Further, the median of coverage percentage for the SMP set is 85.141whereas, for the NSMP set is 33.486, which is significantly lower. These observations indicate that, in general, those journals that have social media plugins integrated in their webpages get higher proportion of their articles covered in social media platforms. It is interesting to find that several very well-recognized journals like IEEE Transactions on Power Systems, Journal of Mathematical Biology, IEEE Transactions on Industrial Electronics etc. have lower overall social media attention to their articles. The absence of social media plugins in their webpages may be one of the primary reasons for this pattern. In order to have a more detailed understanding and clear presentation of the results, the coverage levels were clubbed in 5 groups of 20% each. Figure 1 shows the clubbed coverage levels plotted for journals from both the SMP and NSMP sets, figure on the left for the SMP set and figure on the right for the NSMP set. The x-axis represents the coverage levels and the y-axis represents the number of journals (frequency in that coverage level). Five different bars represent the frequency of each coverage level, in each figure. It is observed that, in the case of SMP set, there are more than 25 journals (which is more than 50% of SMP set) grouped in the range of 80%-100% coverage levels in different social media platforms. Only three journals are having coverage levels less than equal to 20% in the SMP set. In contrast, most of the journals in NSMP set (precisely 27 out of 50 journals) are having coverage levels less than 40%. Only four journals are found to be in a higher coverage level group (80%-100%), which is merely 8% of the whole NSMP set. These observations provide a clear indication that the journals with social media plugin integrated tend to have more social media attention to their his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/ articles. The integration of social media plugins in journal webpages is, thus, observed to have a clearly identifiable advantage in terms of social media coverage. It would also be important here to look at any possible interaction of the above noted patterns with disciplinary variations in social media coverage of articles. Several previous studies (such as Banshal et al., 2019a; 2019b) have shown that articles in some disciplines tend to get higher social media coverage as compared to other disciplines. Articles in disciplines like medical and biological sciences have been found to get higher social media coverage and attention as compared to several other disciplines like Information Science, Mathematics, Engineering etc. We, therefore, look at the disciplinary representation of journals in the two sets and their social media coverage levels. The SMP set comprises of journals from all major disciplines like, Medical Sciences, Biological Sciences, Chemistry, Material Sciences, Mathematics, Earth Sciences, Agricultural Sciences, Environmental Sciences, Engineering, and Multidisciplinary. The NSMP set also has journals representing all major disciplines. If we look at some specific examples, we could understand it more clearly that social media plugin integration is the main differentiating factor across the journals in the same discipline. For example, in SMP set, “Journal of Infectious Diseases”, a journal from Medical Sciences, has coverage percentage of 88.42%, whereas another journal, “Biomed Research International”, from the same discipline, in NSMP set has coverage percentage of 41.31%. Another example could be from the Material Science area, where it is observed that journals, “ACS Nano”, “Materials Today”, “Small”, all in SMP set, get coverage percentages of 75%, 69% and 46%, respectively. On the other hand, a journal “Rapid Prototyping Journal”, again in the Material Science area, in NSMP set has coverage percentage of 4.5%. There are several other examples of similar kind that substantiate the fact that social media plugin presence is a major differentiator in coverage across the journals in the same discipline.
Journals with and without Twitter plugin
After analyzing the overall coverage levels, we analyzed the platform-wise coverage and average mention statistics. The first analysis was for the Twitter data of journals in SMP and NSMP sets. Here, SMP represents the journals that have Twitter plugin integrated and NSMP represents journals that do not have Twitter plugin integrated in their webpages. Tables 3 and 4 present the coverage percentage and tweets per paper for journals in SMP and NSMP sets, respectively. It can be observed here that the mean coverage percentage of articles in Twitter platform for SMP set is 62.3%, which is more than three times higher than the mean coverage percentage of NSMP set, which is 21.71%. Further, the average median values of coverage percentages for the SMP and NSMP sets, for Twitter platform data, are 73.25% and 14.02%, respectively. It can be said that average coverage percentage, for Twitter platform data, for SMP and NSMP sets are significantly different. If we look at individual journals, we observe that in the SMP set, there are 18 journals (out of 50) that have coverage percentage of more than 90%. About 60% journals in the SMP set have coverage percentage higher than 60%. On the other hand, in the NSMP set, only 12 journals (out of 50) are found to have a good coverage percentage. In addition to the coverage percentage, the average tweets per paper was also computed for each journal in the two sets. Average tweets per paper value for a journal is computed by dividing total tweets for articles in that journal by the number of articles. These journal-wise average tweet values are then aggregated for the SMP and NSMP sets. It is observed that the SMP set has mean tweets per paper value as 14.135, whereas the NSMP set has a mean of tweets per paper value as 2.925. The median values of tweets per paper for the SMP and NSMP sets are 6.515 and 2.355, respectively. Some specific examples of journals to look may be Proceedings of the IEEE, IEEE Transactions on Pattern Analysis and Machine his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/
Intelligence, IEEE Transactions on Industrial Electronics, which all, despite being very well-recognized journals, have quite low Twitter coverage and average tweet per paper values being 10.795, 2.504 and 1.267, respectively. Integration of Twitter plugin in journal webpages thus seems to have a clear advantage here. To further understand the differentiation between the whole data of SMP and NSMP sets, a comparative visualization is created with the average tweets per paper values in the two sets. Figure 2 presents the plot of average tweets per paper value for all the journals in the two sets. Here, the x-axis represents a journal and the y-axis represents tweet per paper value for that journal. For observing a clear differentiation, journals on the x-axis are ordered (from left to right) in descending order of their tweet per paper value. The two curves plot average tweets per paper value for different journals in the two sets. It is clearly observed that journals in the SMP set get higher values for tweets per paper as compared to journals in the NSMP set. The curves indicate that a larger number of journals in the SMP set has significantly higher average tweet per paper value as compared to NSMP set. Since we used average tweet per paper value as a differentiating feature across the two sets (classes) of journals, we analyzed the data at a finer granularity of tweets for individual articles. To measure the degree of association between tweets and journal class, a ROC curve is plotted by taking tweets of publications in both sets as the independent variable. Figure 3 shows the ROC curve for the tweet data for articles in journals in the two sets. It is observed that AUC value is 0.730, which shows that a large area is covered under the curve. The large AUC shows a strong association between the tweets and journal classes. A value of AUC = 0.725 in an earlier study by Hassan et al., (2017) used a similar argument. This AUC value thus confirms the association between tweets per paper and journals classes and serves as evidence that average tweets per paper would also have a strong association, and hence can be used to differentiate articles in journals in SMP and NSMP sets. The discipline connection in the case of Twitter platform is also analyzed. We compared the coverage percentage and average tweets per paper for different journals in the two sets. Some interesting example to list would be, journal “Nature Communications” in SMP set, with average tweet per paper value of 24.459, as compared to journal “Complexity” in NSMP set, having average tweet per paper value as 9.444. Both of these journals represent the same disciplinary area. Another example could be journal “Plant, Cell and Environment” in SMP set, with Twitter coverage of 35.455% and average tweet per paper as 4.167, as against journal “British Food Journal” in NSMP set, having Twitter coverage of 12.766% and average tweet per paper as 3.375. Both these journals are from the Agricultural Sciences area. Thus, in the case of Twitter platform data as well, integration of plugin shows a clear differentiation in coverage and mentions per paper, for journals from the same discipline.
Journals with and without Facebook plugin
The Facebook platform data was the next to be analyzed. Tables 5 and 6 present the results for mean coverage percentage and mentions per paper value for the Facebook platform for SMP and NSMP sets, respectively. It is observed that the mean coverage percentage for SMP set is 17.942%, whereas for the NSMP set it is 7.296%. The median of coverage percentage for SMP and NSMP sets are 11.498% and 0.973%, respectively. In terms of Facebook mentions per paper, we have observed that the mean mentions per paper value for SMP set is 1.948and for NSMP set it is 1.001. The median values of mentions per paper for SMP and NSMP sets are 1.462 and 1, respectively. We observe that, in the SMP set, there are 22 journals having his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/ coverage percentage of more than 10%, as against 5 journals in NSMP set. There are 15 journals in NSMP set that have coverage percentage in the Facebook platform as 0. In order to clearly visualize the overall pattern of average Facebook mentions per paper, we create plot with journals on the x-axis and average Facebook mention per paper for that journal on the y-axis. Figure 4 shows the two curves for SMP and NSMP set of journals, with journals on the x-axis arranged in descending order of their average Facebook mention per paper for each journal. A reasonable difference between the average mentions in the two sets can be observed in the plot, with SMP set getting higher values as compared to NSMP set. The peak value for the SMP set is higher than 10 as compared to peak value for the NSMP set around 4. Thus, it is observed that the average mentions per paper value for the journals in the two sets vary, with SMP having overall higher values as compared to the NSMP set. In order to test for the discriminative capacity of average mentions per paper across the two sets of journals, a ROC curve for Facebook platform data is plotted, in figure 5. It is observed that AUC value is 0.555, which is not significantly high. Thus, Facebook mentions per paper has a weak association with the classification of journals in the two sets. The AUC value suggests that average FB mentions are not as encouraging as that of tweets per paper. Relatively low use of Facebook by academicians from several regions of the World could be a reason for this (Hank et al., 2014). The lesser magnitude of differences in Facebook mentions per paper in the two sets shows a similar pattern. The discipline connection has also been analyzed for Facebook data. The journals in SMP and NSMP sets are compared in terms of coverage levels and average mentions for several disciplines. Though the magnitude of differences is not high in case of the Facebook platform, still some positive advantage is seen of having Facebook plugin integrated in journals. For example, the journal “Cell”, in SMP set, has a coverage percentage of 60.98% and average mentions per paper as 4.345, as against the journal “Parasitology Research”, in NSMP set, with coverage percentage as 6.767% and average mentions per paper as 1.556. There are several other examples of a slight advantage of integrating Facebook plugin in journals, of the same discipline, in terms of coverage percentage and mentions per paper. Thus, perusal of all the results for Facebook platform show that in the case of this platform too, there is an observed advantage of integrating plugin in journal webpages on their coverage and mentions per paper value. However, here the magnitude of differences is not very high between the journals in the two sets.
Journals with and without Mendeley plugin
The third platform for which data was analyzed was Mendeley. In this case, unlike the other two platforms, we found only 22 journals having Mendeley plugin integrated. Thus, in this case, the NSMP set comprises of 78 journals. Tables 7 and 8 show the results for mean coverage percentage and mentions per paper for SMP and NSMP sets, respectively. It is observed that the mean coverage percentage for SMP and NSMP sets are 84.927% and 36.256%. Similarly, the median values for coverage percentage for SMP and NSMP sets are 94.252% and 32.36%, respectively. Thus, a significant value of difference in coverage percentage is seen in journals and SMP and NSMP sets. The mean and median values for mentions per paper was also computed for both sets. It is observed that for SMP set the mean of mentions per paper is 49.656 and the median is 50.971, whereas for NSMP set mean and median of mentions per paper are37.002 and24.899. Though, the mean value of mentions per paper of the NSMP set is close to the SMP set, the median value of the SMP set is higher than two times of the NSMP set. his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/
The spread of average reads per paper for the journals across the two sets (SMP & NSMP) has been plotted in figure 6, for a better visualization. Here too, the x-axis has journals arranged in descending order of Mendeley reads per paper and y-axis value represents the Mendeley reads per paper for that journal. The two curves show the patterns for SMP and NSMP sets. The difference between the average reads in the SMP and NSMP sets is quite high as compared to the variation in Twitter and Facebook platforms. Though the highest value from both the sets lies around 200 but the overall difference between the averages is evidently noticeable in trend lines from the figure. The tail trend of the NSMP set is quite lower than the SMP set. The larger size of the NSMP set could be a reason for a relatively much higher difference in this case. Nevertheless, it is observed that journals with Mendeley plugin get much higher reads per paper as compared to journals that do not have such plugin. The discriminative capacity of average Mendeley reads is also analyzed through the ROC plot by looking at article level read statistics. Figure 7 shows the ROC curve, representing the relationship between journal class and Mendeley reads. The AUC value in this case is 0.673, which could be taken as an acceptable value for the discriminative capacity of Mendeley reads between the two classes. This value thus indicates the validity of representing the average reads as the differentiating indicator between SMP and NSMP journals and confirms its use in the analysis. The coverage and attention levels of the Mendeley platform have also been analyzed with respect to discipline connection. In the case of Mendeley platform data too, there are several examples of journals in the same discipline having differential coverage and average reads per paper value owing to Mendeley plugin integration. One example could be a journal, “Journal of The American Statistical Association”, in the SMP set, that has coverage of 95.031% and average reads per paper as 24.353. Another journal “Acta Mathematica Sinica”, in NSMP set from the similar area, has 5.172 % coverage and 6.5 average reads per paper. The Medical Sciences journal “PLoS Neglected Tropical Diseases”, in the SMP set has coverage 97.603% and average reads per paper 55.24 as against to the journal “Disease Markers”, in the NSMP set with coverage 41.615% and the average reads per paper as 22.687. Thus, the plugin advantage of social media attention is observed in Mendeley platform too, as observed in journals from the same discipline in the two sets. The difference in magnitudes here, however, is not as significant as Twitter.
Conclusion
The paper tried to answer the question, whether the presence of social media plugin in journal webpages result in higher social media visibility of its articles. Two data samples of 50 journals each, with and without social media plugin, are studied and the social media activity around their publications is captured. Social media data of a total of 99,749 articles, from these 100 journals, were obtained and analyzed for the purpose. The journals in the data sample represent different disciplines as well as different publishers. The coverage and mentions per paper values in three platforms- Twitter, Facebook and Mendeley- are analyzed. The ROC curves are also plotted and AUC computed for mentions per paper data for various platforms to understand the discriminative power of mentions per paper values for the two journal sets. Results show a clearly identifiable differentiation between social media coverage and mention per paper for articles published in a journal with and without social media plugins. A journal that has a plugin for a social platform integrated in its website is found to attract higher coverage and attention to its articles in that social media platform. This is found to be valid for all the three social media platforms analyzed, though the degree of differentiation varies across his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/ the platforms. The impact of the presence of plugin is seen most clearly and notably for the Twitter platform. The journals that have Twitter plugin integrated in their websites get much higher coverage and attention of their articles in Twitter. Mendeley platform also shows differentiation in coverage and mentions per paper between the journals, with and without Mendeley plugins. The Facebook plugin integration in journals also shows a positive impact, though the magnitude of difference in impact on social media coverage and attention is lesser as compared to other platforms. The ROC plots made and the AUC values computed for article-level mentions per paper data, for the three platforms, show that average mentions per paper is a suitable discriminator of the two sets of journals, those with the plugin and those without them. Twitter mention data has the highest value of AUC followed by Mendeley. Facebook platform mention data shows relatively lower AUC value. The values indicate higher differentiation in mentions per paper values in the Twitter platform as compared to Facebook. The set of journals analyzed not only represented different publishers but also different disciplines. The social media data of articles is, thus, a suitable representation of various disciplines. Given the fact that some disciplines are known to attract higher social media attention, it was important to see that the positive differentiation of the presence of social media plugin on social media coverage and attention holds for journals in different disciplines. The results show several examples of journals from the same disciplinary area, in SMP and NSMP sets, where journals in the SMP set get much higher social media coverage and attention of their articles as compared to journals in NSMP set. This confirms that the impact of the presence of social media plugin in journal webpages is independent of discipline. The different results for overall coverage, mean and median values of coverage percentages, mentions per paper values and the AUC values computed for journals in two sets for different platforms confirm that presence of social media plugin in journal webpages give higher social media coverage and attention to articles published in that journal. The results, therefore, provide a positive answer to the research question proposed through conclusive evidence from three social platforms. It would be important to try to understand possible reasons for these patterns observed and to look at what practical implications it may have for scholarly publishing. It appears that the presence of social media plugin on journal webpages acts as a kind of facilitator for social media sharing and mentioning. Perhaps, authors and readers prefer to click the readily available plugin links on the webpages to either share the article in their profiles or to write their comments/ opinions about the article. These clicks and shares may be originating from different kinds of browsing sessions. First, authors of articles in the journal may find it easier and more convenient to use the plugins available at the journal webpage to disseminate and share their articles. Secondly, the readers of a journal may also find it useful to use the plugin links to share new findings reported in the articles with their peers and students. Both these activities result in higher social shares & mentions of articles published in journals that provide readily available sharing/ commenting mechanism in the form of plugins. However, journals that do not provide the plugins integrated in their webpages are at a disadvantage in terms of ease and convenience of sharing in social media platforms. These results have very important practical implications for journal publishers. Results indicate that publishing houses should actively provide social media plugin integration in their journal webpages to increase the social media visibility (altmetric impact) of their articles. Given the fact that altmetric values correlate well with citation counts of articles, this indirectly implies that integration of social media plugins in journal webpage may also indirectly facilitate in higher citation impact of articles in the journal. Journals that do not integrate the social media his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/ plugin may continue to be at a disadvantage in respect to social media sharing/ mentions of their articles.
Acknowledgements
The authors would like to acknowledge the support of Stacy Konkiel, Director of Research Relations at Digital Science, for providing access to Altmetric.com data.
References
Banshal, S. K., Singh, V. K., Kaderye, G., Muhuri, P. K., & Sánchez, B. P. (2018). An altmetric analysis of scholarly articles from India.
Journal of Intelligent & Fuzzy Systems , (5), 3111–3118. https://doi.org/10.3233/JIFS-169495 Banshal, S. K., Singh, V. K., Muhuri, P. K., & Mayr, P. (2019a). How much research output from India gets social media attention? Current Science , (5), 753–760. Banshal, S. K., Singh, V. K., Muhuri, P. K., & Mayr, P. (2019b). Disciplinary Variations in Altmetric Coverage of Scholarly Articles. In 17th INTERNATIONAL CONFERENCE ON SCIENTOMETRICS & INFORMETRICS (ISSI) , 1870–1881. Bar-ilan, J. (2014). Astrophysics publications on arXiv , Scopus and Mendeley : a case study.
Scientometrics , (1), 217–225. https://doi.org/10.1007/s11192-013-1215-1 Chen, K., Tang, M., Wang, C., & Hsiang, J. (2015). Exploring alternative metrics of scholarly performance in the social sciences and humanities in Taiwan. Scientometrics , (1), 97–112. https://doi.org/10.1007/s11192-014-1420-6 Costas, R., Zahedi, Z., & Wouters, P. (2015a). Do “altmetrics” correlate with citations? Extensive comparison of altmetric indicators with citations from a multidisciplinary perspective. Journal of the Association for Information Science and Technology , (10), 2003–2019. https://doi.org/10.1002/asi.23309 Costas, R., Zahedi, Z., & Wouters, P. (2015b). The thematic orientation of publications mentioned on social media: Large-scale disciplinary comparison of social media metrics with citations. Aslib Journal of Information Management , (3), 260–288. https://doi.org/10.1108/AJIM-12-2014-0173 Davis, P. M., Lewenstein, B. V, Simon, D. H., Booth, J. G., & Connolly, M. J. L. (2008). Open access publishing, article downloads, and citations: randomised controlled trial. BMJ (Clinical Research Ed.) , , a568. https://doi.org/10.1136/bmj.a568 Didegah, F., Bowman, T. D., & Holmberg, K. (2018). On the Differences Between Citations and Altmetrics : An Investigation of Factors Driving Altmetrics Versus Citations for Finnish Articles. JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY , (6), 832–843. https://doi.org/10.1002/asi.23934 Eysenbach, G. (2011). Can tweets predict citations? Metrics of social impact based on Twitter and correlation with traditional metrics of scientific impact. Journal of Medical Internet Research , (4). https://doi.org/10.2196/jmir.2012 Gargouri, Y., Hajjem, C., Larivière, V., Gingras, Y., Carr, L., Brody, T., & Harnad, S. (2010). Self-selected or mandated, open access increases citation impact for higher quality research. PloS One , (10), e13636. Gruzd, A., & Goertzen, M. (2013). Wired academia: Why social science scholars are using his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/ social media. Proceedings of the Annual Hawaii International Conference on System Sciences , 3332–3341. https://doi.org/10.1109/HICSS.2013.614 Hammarfelt, B. (2014). Using altmetrics for assessing research impact in the humanities.
Scientometrics , (2), 1419–1430. https://doi.org/10.1007/s11192-014-1261-3 Hank, C., Sugimoto, C. R., Tsou, A., & Pomerantz, J. (2014). Faculty and student interactions via Facebook: Policies, preferences, and practices. It-Information Technology , (5), 216–223. Hassan, S., Mubashir, I., Gillani, U., Aljohani, N. R., Bowman, T. D., & Didegah, F. (2017). Measuring social media activity of scientific literature : an exhaustive comparison of scopus and novel altmetrics big data. Scientometrics , (2), 1037–1057. https://doi.org/10.1007/s11192-017-2512-x Haustein, S., Costas, R., & Larivière, V. (2015). Characterizing social media metrics of scholarly papers: The effect of document properties and collaboration patterns. PLoS ONE , (3), 1–21. https://doi.org/10.1371/journal.pone.0120495 Haustein, S., Peters, I., Sugimoto, C. R., Thelwall, M., & Larivière, V. (2014). Tweeting biomedicine: An analysis of tweets and citations in the biomedical literature. Journal of the Association for Information Science and Technology , (4), 656–669. Holmberg, K., Hedman, J., Bowman, T. D., Didegah, F., & Laakso, M. (2020). Do articles in open access journals have more frequent altmetric activity than articles in subscription-based journals? An investigation of the research output of Finnish universities. Scientometrics , (1), 645–659. https://doi.org/10.1007/s11192-019-03301-x Htoo, T. H. H., & Na, J.-C. (2017). Disciplinary Differences in Altmetrics for Social Sciences. Online Information Review , (2), 235–251. Maggio, L. A., Leroux, T. C., Meyer, H. S., & Artino, A. R. (2018). Perspectives on Medical Education , (4), 239–247. https://doi.org/10.1007/s40037-018-0438-5 Norris, M., Oppenheim, C., & Rowland, F. (2008). The citation advantage of open-access articles. Journal of the American Society for Information Science and Technology , , 1963–1972. https://doi.org/10.1002/asi.20898 Ortega, J. L. (2015). Disciplinary differences in the use of academic social networking sites. Online Information Review , (4), 520–536. Poplašen, L. M., & Grgić, I. H. (2017). Altmetric and Bibliometric Scores: Does Open Access Matter? QQML Journal , (2), 451–460. Priem, J. (2014). Altmetrics (Chapter from Beyond Bibliometrics: Harnessing Multidimensional Indicators of Scholarly Impact). In Beyond Bibliometrics: Harnessing Multidimensional Indicators of Scholarly Impact . Priem, J., & Hemminger, B. H. (2010). Scientometrics 2.0: New metrics of scholarly impact on the social Web.
First Monday , (7), 1–19. Priem, J., Taraborelli, D., Groth, P., & Neylon, C. (2010). altmetrics : a manifesto . 1–77. Shema, H., Bar-IIan, J., & Thelwall, M. (2014). Do Blog Citations Correlate With a Higher Number of Future Citations? Research Blogs as a Potential Source for Alternative Metrics. JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY , (5), 1018–1027. https://doi.org/10.1002/asi his is a pre-print of an article published in [Scientometrics]. The final authenticated version is available online at: https://doi.org/ Sotudeh, H., Mazarei, Z., & Mirzabeigi, M. (2015). CiteULike bookmarks are correlated to citations at journal and author levels in library and information science.
Scientometrics , (3), 2237–2248. https://doi.org/10.1007/s11192-015-1745-9 Thelwall, M. (2018). Early Mendeley readers correlate with later citation counts. Scientometrics , (3), 1231–1240. https://doi.org/10.1007/s11192-018-2715-9 Thelwall, M., & Kousha, K. (2017). ResearchGate articles: Age, discipline, audience size, and impact. Journal of the Association for Information Science and Technology , (2), 468–479. https://doi.org/10.1002/asi.23675 Thelwall, M., & Nevill, T. (2018). Could scientists use Altmetric . com scores to predict longer term citation counts ? Journal of Informetrics , (1), 237–248. Vogl, S., Scherndl, T., & Ku, A. (2018). Scientometrics , (3), 1253–1269. https://doi.org/10.1007/s11192-018-2727-5 Wang, X., Liu, C., Mao, W., & Fang, Z. (2015). The open access advantage considering citation , article usage and social media attention. Scientometrics , (2), 555–564. https://doi.org/10.1007/s11192-015-1547-0 Williams, A. E., & Woodacre, M. A. (2016). The possibilities and perils of academic social networking sites. Online Information Review , (2), 282–294. https://doi.org/10.1108/OIR-10-2015-0327 Zhang, L., & Wang, J. (2018). Why highly cited articles are not highly tweeted? A biology case. Scientometrics , (1), 495–509. https://doi.org/10.1007/s11192-018-2876-6 IGURES
Figure 1: Coverage Levels of SMP & NSMP sets of journals Figure 2: Comparison of average Tweets per paper for journals in SMP & NSMP sets Figure 3: ROC curve for average Tweets to classify papers in SMP & NSMP sets Figure 4: Comparison of average FB mentions per paper for journals in SMP & NSMP sets Figure 5: ROC Curve for average FB mentions to classify papers in SMP & NSMP sets Figure 6: Comparison of average Mendeley reads per paper for journals in SMP & NSMP sets Figure 7: ROC Curve for average Mendeley reads to classify papers in SMP & NSMP sets
ABLES
Table1: Publisher and Coverage Summary for journals with social media plugin S. No. Journal Name Publisher WoS Records Records with DOI Records found in altmetric.com Coverage Level % Twitter FB Mendeley
1 Acs Nano Acs Publications 1,285 1,285 975 75.875 YES YES YES 2 Advanced Energy Materials Wiley 420 419 94 22.434 YES YES NO 3 Advanced Functional Materials Wiley 881 881 322 36.549 YES YES NO 4 Advanced Materials Wiley 1,172 1,169 70 5.988 YES YES NO 5 American Journal of Sports Medicine Sage 438 438 388 88.584 YES YES YES 6 Annals of Botany Oxford University Press 217 209 208 99.522 YES YES YES 7 Applied Catalysis B Environmental Elsevier 789 788 348 44.162 YES YES NO 8 Bioinformatics Oxford University Press 812 811 809 99.753 YES YES YES 9 Biomaterials Elsevier 651 651 451 69.278 YES YES NO 10 Biophysical Journal Cell Press 3,814 3,644 874 23.985 YES YES NO 11 Biosensors & Bioelectronics Elsevier 1,015 1,014 699 68.935 YES YES NO 12 Briefings in Bioinformatics Oxford University Press 93 93 93 100 YES YES YES 13 Cell Reports Cell Press 1,101 1,100 1,081 98.273 YES YES NO 14 Cell Cell Press 690 661 648 98.033 YES YES NO 15 Chinese Journal of Catalysis Elsevier 239 239 9 3.766 YES YES NO 16 Clinical Infectious Diseases Oxford University Press 839 777 628 80.824 YES YES YES 17 Current Biology Cell Press 847 846 825 97.518 YES YES NO 8 Database The Journal of Biological Databases And Curation Oxford University Press 166 164 139 84.756 YES YES YES 19 European Journal of Sport Science Taylor & Francis 155 155 153 98.71 YES YES YES 20 Expert Review of Medical Devices Taylor & Francis 116 116 116 100 YES YES YES 21 International Journal of Production Research Taylor & Francis 460 460 118 25.652 YES YES YES 22 International Journal of Remote Sensing Taylor & Francis 334 334 86 25.749 YES YES YES 23 Journal of Catalysis Elsevier 358 358 87 24.302 YES YES NO 24 Journal of Experimental Botany Oxford University Press 514 514 487 94.747 YES YES YES 25 Journal of Infectious Diseases Oxford University Press 691 691 611 88.423 YES YES YES 26 Journal of Power Sources Elsevier 1,651 1,651 320 19.382 YES YES NO 27 Journal of Sports Sciences Taylor & Francis 298 297 292 98.316 YES YES YES 28 Journal of The American Statistical Association Taylor & Francis 161 161 159 98.758 YES YES YES 29 Materials Today Elsevier 183 96 66 68.75 YES YES NO 30 Molecular Ecology Wiley 433 433 424 97.921 YES YES NO 31 Nano Letters Acs Publications 1,195 1,194 993 83.166 YES YES YES 32 Nanotoxicology Taylor & Francis 150 150 93 62 YES YES YES 33 Nature Communications Nature 3,687 3,685 3,452 93.677 YES YES NO 34 Nature Materials Nature 291 277 242 87.365 YES YES NO 35 Nature Nature 2,807 2,397 2,390 99.708 YES YES NO 36 New Phytologist Wiley 538 530 524 98.868 YES YES NO 37 NPG Asia Materials Nature 104 104 98 94.231 YES YES NO 38 Nucleic Acids Oxford 1,298 1,298 1,226 94.453 YES YES YES
Table 2: Publisher and Coverage Summary for journals without social media plugin S. No. Journal name Publisher WoS Records Records with DOI Records found in altmetric.com Coverage Level % Twitter FB Mendeley
1 Acta Mathematica Sinica Springer 116 116 8 6.897 NO NO NO 2 Advances in Mathematical Physics Hindawi 122 122 17 13.934 NO NO NO 3 Aircraft Engineering and Aerospace Technology Emerald 90 90 5 5.556 NO NO NO 4 Analytical & Bioanalytical Chemistry Springer 837 837 539 64.397 NO NO NO 5 Annals of Biomedical Engineering Springer 306 305 200 65.574 NO NO NO Research University Press 39 Plant Biotechnology Journal Wiley 203 203 174 85.714 YES YES NO 40 Plant, Cell And Environment Wiley 221 220 97 44.091 YES YES NO 41 PLoS Computational Biology PLoS 582 582 580 99.656 YES YES YES 42 PLoS Neglected Tropical Diseases PLoS 876 876 856 97.717 YES YES YES 43 PLoS One PLoS 23,020 23,019 19,687 85.525 YES YES YES 44 PLoS Pathogens PLoS 685 684 683 99.854 YES YES YES 45 Progress in Materials Science Elsevier 49 49 24 48.98 YES YES NO 46 Progress in Photovoltaics Wiley 143 143 48 33.566 YES YES NO 47 Progress in Polymer Science Elsevier 49 49 22 44.898 YES YES NO 48 Scientific Reports Nature 21,063 21,057 14,968 71.083 YES YES NO 49 Small Wiley 660 659 300 45.524 YES YES NO 50 Statistical Methods in Medical Research Sage 189 189 130 68.783 YES YES NO
Mean 71.556
Median 85.141
Annals of Operations Research Springer 312 312 72 23.077 NO NO NO 7 Applied Microbiology & Biotechnology Springer 887 887 523 58.963 NO NO NO 8 Autonomous Robots Springer 78 78 78 100 NO NO NO 9 Biochemistry-Moscow Springer 171 171 46 26.901 NO NO NO 10 Biomed Research International Hindawi 1853 1852 766 41.361 NO NO NO 11 British Food Journal Emerald 188 188 66 35.106 NO NO NO 12 Bulletin of Mathematical Biology Springer 95 95 61 64.211 NO NO NO 13 Canadian Respiratory Journal Hindawi 95 95 50 52.632 NO NO NO 14 Chinese Annals of Mathematics Springer 73 73 1 1.37 NO NO NO 15 Complexity Hindawi 194 194 56 28.866 NO NO NO 16 Disease Markers Hindawi 161 161 67 41.615 NO NO NO 17 Empirical Software Engineering Springer 73 73 68 93.151 NO NO NO 18 IEEE Access IEEE 820 815 23 2.822 NO NO NO 19 IEEE Communications Letters IEEE 635 635 127 20 NO NO NO 20 IEEE Journal on Selected Areas in Communications IEEE 301 301 109 36.213 NO NO NO 21 IEEE Transactions on Antennas and Propagation IEEE 677 677 150 22.157 NO NO NO 22 IEEE Transactions on Biomedical Engineering IEEE 269 269 156 57.993 NO NO NO 23 IEEE Transactions on Communications IEEE 405 405 143 35.309 NO NO NO 24 IEEE Transactions on Geoscience and Remote Sensing IEEE 573 573 147 25.654 NO NO NO 25 IEEE Transactions on Image Processing IEEE 446 446 183 41.031 NO NO NO 26 IEEE Transactions on IEEE 766 765 157 20.523 NO NO NO ndustrial Electronics 27 IEEE Transactions on Mobile Computing IEEE 225 224 94 41.964 NO NO NO 28 IEEE Transactions on Neural Networks and Learning Systems IEEE 228 228 92 40.351 NO NO NO 29 IEEE Transactions on Pattern Analysis and Machine Intelligence IEEE 193 193 142 73.575 NO NO NO 30 IEEE Transactions on Power Electronics IEEE 736 736 146 19.837 NO NO NO 31 IEEE Transactions on Power Systems IEEE 529 528 150 28.409 NO NO NO 32 IEEE Transactions on Signal Processing IEEE 486 486 249 51.235 NO NO NO 33 IEEE Transactions on Visualization and Computer Graphics IEEE 245 240 82 34.167 NO NO NO 34 IEEE Transactions on Wireless Communications IEEE 631 631 207 32.805 NO NO NO 35 Industrial Management & Data Systems Emerald 96 96 30 31.25 NO NO NO 36 International Journal of Advanced Manufacturing Technology Springer 1420 1420 317 22.324 NO NO NO 37 International Journal of Numerical Methods for Heat & Fluid Flow Emerald 132 132 11 8.333 NO NO NO 38 Journal of Lightwave Technology IEEE 691 691 158 22.865 NO NO NO 39 Journal of Mathematical Biology Springer 133 133 72 54.135 NO NO NO 40 Kybernetes Emerald 101 101 7 6.931 NO NO NO 41 Multimedia Tools & Applications Springer 876 876 105 11.986 NO NO NO 42 Online Information Review Emerald 68 68 28 41.176 NO NO NO 43 Parasitology Research Springer 532 532 439 82.519 NO NO NO 4 Proceedings of The IEEE IEEE 159 155 95 61.29 NO NO NO 45 Rapid Prototyping Journal Emerald 90 90 9 10 NO NO NO 46 Scanning Hindawi 102 102 21 20.588 NO NO NO 47 Security and Communication Networks Hindawi 475 475 69 14.526 NO NO NO 48 Shock and Vibration Hindawi 390 390 46 11.795 NO NO NO 49 Springerplus Springer 2115 2115 850 40.189 NO NO NO 50 Surgical Endoscopy & Other Interventional Techniques Springer 753 752 649 86.303 NO NO NO
Mean 36.677
Median 33.486
Table 3: Mention statistics for journals with Twitter plugin S. No. Journal Name Records with DOI Tweeted Coverage % Twitter Mentions Avg. Twitter Mentions
1 Acs Nano 1,285 793 61.712 3,941 4.97 2 Advanced Energy Materials 419 48 11.456 68 1.417 3 Advanced Functional Materials 881 123 13.961 327 2.659 4 Advanced Materials 1,169 41 3.507 123 3 5 American Journal of Sports Medicine 438 378 86.301 9,565 25.304 6 Annals of Botany 209 199 95.215 2,866 14.402 7 Applied Catalysis B Environmental 788 305 38.706 741 2.43 8 Bioinformatics 811 808 99.63 9,454 11.7 9 Biomaterials 651 326 50.077 831 2.549 10 Biophysical Journal 3,644 468 12.843 2,347 5.015 11 Biosensors & Bioelectronics 1,014 635 62.623 1,471 2.317 12 Briefings in Bioinformatics 93 93 100 1,159 12.462 13 Cell Reports 1,100 1,071 97.364 16,028 14.965 14 Cell 661 642 97.126 37,557 58.5 15 Chinese Journal of Catalysis 239 2 0.837 2 1 16 Clinical Infectious Diseases 777 535 68.855 10,686 19.974 17 Current Biology 846 796 94.09 31,275 39.29 18 Database The Journal of Biological 164 123 75 887 7.211 atabases And Curation 19 European Journal of Sport Science 155 153 98.71 4,114 26.889 20 Expert Review of Medical Devices 116 116 100 203 1.75 21 International Journal of Production Research 460 23 5 48 2.087 22 International Journal of Remote Sensing 334 30 8.982 206 6.867 23 Journal of Catalysis 358 38 10.615 93 2.447 24 Journal of Experimental Botany 514 467 90.856 2,476 5.302 25 Journal of Infectious Diseases 691 515 74.53 4,398 8.54 26 Journal of Power Sources 1,651 93 5.633 221 2.376 27 Journal of Sports Sciences 297 290 97.643 6,977 24.059 28 Journal of The American Statistical Association 161 158 98.137 605 3.829 29 Materials Today 96 21 21.875 109 5.19 30 Molecular Ecology 433 415 95.843 5,541 13.352 31 Nano Letters 1,194 775 64.908 6,287 8.112 32 Nanotoxicology 150 62 41.333 138 2.226 33 Nature Communications 3,685 3,132 84.993 76,607 24.459 34 Nature Materials 277 222 80.144 3,199 14.41 35 Nature 2,397 2,386 99.541 487,961 204.51 36 New Phytologist 530 522 98.491 7,172 13.739 37 NPG Asia Materials 104 98 94.231 242 2.469 38 Nucleic Acids Research 1,298 1,144 88.136 8,157 7.13 39 Plant Biotechnology Journal 203 148 72.906 912 6.162 40 Plant, Cell And Environment 220 78 35.455 325 4.167 41 PLoS Computational Biology 582 576 98.969 14,629 25.398 42 PLoS Neglected Tropical Diseases 876 843 96.233 10,954 12.994 43 PLoS One 23,019 16,942 73.6 160,665 9.483 44 PLoS Pathogens 684 676 98.83 9,968 14.746 45 Progress in Materials Science 49 18 36.735 86 4.778 46 Progress in Photovoltaics 143 18 12.587 37 2.056 47 Progress in Polymer Science 49 10 20.408 29 2.9 48 Scientific Reports 21,057 12,442 59.087 124,815 10.032 49 Small 659 187 28.376 378 2.021 50 Statistical Methods in Medical Research 189 100 52.91 309 3.09
Mean
1 Acta Mathematica Sinica 116 6 5.172 31 5.167 2 Advances in Mathematical Physics 122 15 12.295 40 2.667 3 Aircraft Engineering and Aerospace Technology 90 1 1.111 2 2 4 Analytical & Bioanalytical Chemistry 837 486 58.065 1,116 2.296 5 Annals of Biomedical Engineering 305 118 38.689 310 2.627 6 Annals of Operations Research 312 25 8.013 64 2.56 7 Applied Microbiology & Biotechnology 887 433 48.816 970 2.24 8 Autonomous Robots 78 78 100 183 2.346 9 Biochemistry-Moscow 171 31 18.129 73 2.355 10 Biomed Research International 1,852 518 27.97 1,378 2.66 11 British Food Journal 188 24 12.766 81 3.375 12 Bulletin of Mathematical Biology 95 45 47.368 163 3.622 13 Canadian Respiratory Journal 95 34 35.789 72 2.118 14 Chinese Annals of Mathematics 73 0 0 0 15 Complexity 194 45 23.196 425 9.444 16 Disease Markers 161 45 27.95 73 1.622 17 Empirical Software Engineering 73 22 30.137 132 6 18 IEEE Access 815 9 1.104 30 3.333 19 IEEE Communications Letters 635 43 6.772 77 1.791 20 IEEE Journal on Selected Areas in Communications 301 33 10.963 72 2.182 21 IEEE Transactions on Antennas and Propagation 677 35 5.17 52 1.486 22 IEEE Transactions on Biomedical Engineering 269 75 27.881 271 3.613 3 IEEE Transactions on Communications 405 54 13.333 113 2.093 24 IEEE Transactions on Geoscience and Remote Sensing 573 60 10.471 115 1.917 25 IEEE Transactions on Image Processing 446 82 18.386 272 3.317 26 IEEE Transactions on Industrial Electronics 765 15 1.961 19 1.267 27 IEEE Transactions on Mobile Computing 224 10 4.464 43 4.3 28 IEEE Transactions on Neural Networks and Learning Systems 228 46 20.175 96 2.087 29 IEEE Transactions on Pattern Analysis and Machine Intelligence 193 119 61.658 298 2.504 30 IEEE Transactions on Power Electronics 736 18 2.446 36 2 31 IEEE Transactions on Power Systems 528 39 7.386 120 3.077 32 IEEE Transactions on Signal Processing 486 184 37.86 567 3.082 33 IEEE Transactions on Visualization and Computer Graphics 240 46 19.167 136 2.957 34 IEEE Transactions on Wireless Communications 631 94 14.897 193 2.053 35 Industrial Management & Data Systems 96 10 10.417 22 2.2 36 International Journal of Advanced Manufacturing Technology 1,420 22 1.549 33 1.5 37 International Journal of Numerical Methods for Heat & Fluid Flow 132 1 0.758 1 1 38 Journal of Lightwave Technology 691 59 8.538 142 2.407 39 Journal of Mathematical Biology 133 50 37.594 117 2.34 40 Kybernetes 101 5 4.95 15 3 41 Multimedia Tools & 876 41 4.68 84 2.049 pplications 42 Online Information Review 68 25 36.765 123 4.92 43 Parasitology Research 532 429 80.639 1092 2.545 44 Proceedings of The IEEE 155 39 25.161 421 10.795 45 Rapid Prototyping Journal 90 4 4.444 7 1.75 46 Scanning 102 15 14.706 34 2.267 47 Security and Communication Networks 475 26 5.474 34 1.308 48 Shock and Vibration 390 5 1.282 6 1.2 49 Springerplus 2,115 601 28.416 2345 3.902 50 Surgical Endoscopy & Other Interventional Techniques 752 457 60.771 1823 3.989
Mean 21.714 2.925 Median 14.02 2.355 Table 5: Mention statistics for journals with Facebook plugin S. No. Journal Name Records with DOI Found in FB Coverage % FB Mentions Avg. FB Mentions
1 Acs Nano 1,285 104 8.093 168 1.615 2 Advanced Energy Materials 419 5 1.193 5 1 3 Advanced Functional Materials 881 19 2.157 23 1.211 4 Advanced Materials 1,169 7 0.599 10 1.429 5 American Journal of Sports Medicine 438 216 49.315 929 4.301 6 Annals of Botany 209 126 60.287 255 2.024 7 Applied Catalysis B Environmental 788 9 1.142 11 1.222 8 Bioinformatics 811 153 18.866 183 1.196 9 Biomaterials 651 56 8.602 78 1.393 10 Biophysical Journal 3,644 36 0.988 49 1.361 11 Biosensors & Bioelectronics 1,014 22 2.17 25 1.136 12 Briefings in Bioinformatics 93 14 15.054 18 1.286 13 Cell Reports 1,100 424 38.545 939 2.215 14 Cell 661 403 60.968 1,751 4.345 15 Chinese Journal of Catalysis 239 0 0 0 0 16 Clinical Infectious Diseases 777 185 23.81 495 2.676 17 Current Biology 846 325 38.416 1,316 4.049 8 Database The Journal of Biological Databases And Curation 164 20 12.195 26 1.3 19 European Journal of Sport Science 155 73 47.097 290 3.973 20 Expert Review of Medical Devices 116 7 6.034 8 1.143 21 International Journal of Production Research 460 3 0.652 3 1 22 International Journal of Remote Sensing 334 17 5.09 20 1.176 23 Journal of Catalysis 358 4 1.117 4 1 24 Journal of Experimental Botany 514 93 18.093 139 1.495 25 Journal of Infectious Diseases 691 136 19.682 255 1.875 26 Journal of Power Sources 1,651 18 1.09 21 1.167 27 Journal of Sports Sciences 297 89 29.966 185 2.079 28 Journal of The American Statistical Association 161 2 1.242 2 1 29 Materials Today 96 11 11.458 15 1.364 30 Molecular Ecology 433 96 22.171 204 2.125 31 Nano Letters 1,194 74 6.198 149 2.014 32 Nanotoxicology 150 2 1.333 3 1.5 33 Nature Communications 3,685 1,068 28.982 3,356 3.142 34 Nature Materials 277 65 23.466 189 2.908 35 Nature 2,397 1,936 80.768 19,927 10.293 36 New Phytologist 530 218 41.132 452 2.073 37 NPG Asia Materials 104 12 11.538 18 1.5 38 Nucleic Acids Research 1,298 134 10.324 186 1.388 39 Plant Biotechnology Journal 203 65 32.02 129 1.985 40 Plant, Cell And Environment 220 16 7.273 21 1.312 41 PLoS Computational Biology 582 164 28.179 324 1.976 42 PLoS Neglected Tropical Diseases 876 340 38.813 700 2.059 43 PLoS One 23,019 3,901 16.947 9,231 2.366 44 PLoS Pathogens 684 210 30.702 418 1.99 45 Progress in Materials Science 49 1 2.041 1 1 46 Progress in Photovoltaics 143 6 4.196 6 1 47 Progress in Polymer Science 49 5 10.204 5 1 48 Scientific Reports 21,057 2,540 12.062 6,596 2.597 9 Small 659 25 3.794 29 1.16 50 Statistical Methods in Medical Research 189 2 1.058 2 1
Mean 17.942 1.948 Median 11.498 1.462 Table 6: Mention statistics for journals without Facebook plugin S. No. Journal Name Records with DOI Found in FB Coverage % FB Mentions Avg. FB Mentions
1 Acta Mathematica Sinica 116 0 0 0 0 2 Advances in Mathematical Physics 122 2 1.639 2 1 3 Aircraft Engineering and Aerospace Technology 90 0 0 0 0 4 Analytical & Bioanalytical Chemistry 837 26 3.106 38 1.462 5 Annals of Biomedical Engineering 305 17 5.574 26 1.529 6 Annals of Operations Research 312 2 0.641 2 1 7 Applied Microbiology & Biotechnology 887 26 2.931 48 1.846 8 Autonomous Robots 78 65 83.333 65 1 9 Biochemistry-Moscow 171 3 1.754 3 1 10 Biomed Research International 1,852 121 6.533 215 1.777 11 British Food Journal 188 9 4.787 11 1.222 12 Bulletin of Mathematical Biology 95 15 15.789 18 1.2 13 Canadian Respiratory Journal 95 6 6.316 6 1 14 Chinese Annals of Mathematics 73 0 0 0 0 15 Complexity 194 17 8.763 33 1.941 16 Disease Markers 161 7 4.348 7 1 17 Empirical Software Engineering 73 67 91.781 82 1.224 18 IEEE Access 815 1 0.123 1 1 19 IEEE Communications Letters 635 0 0 0 0 20 IEEE Journal on Selected Areas in Communications 301 0 0 0 0 21 IEEE Transactions on Antennas and Propagation 677 5 0.739 6 1.2 22 IEEE Transactions on Biomedical Engineering 269 9 3.346 10 1.111 23 IEEE Transactions on Communications 405 0 0 0 0 24 IEEE Transactions on Geoscience and 573 5 0.873 6 1.2 emote Sensing 25 IEEE Transactions on Image Processing 446 1 0.224 2 2 26 IEEE Transactions on Industrial Electronics 765 0 0 0 0 27 IEEE Transactions on Mobile Computing 224 2 0.893 2 1 28 IEEE Transactions on Neural Networks and Learning Systems 228 7 3.07 8 1.143 29 IEEE Transactions on Pattern Analysis and Machine Intelligence 193 3 1.554 5 1.667 30 IEEE Transactions on Power Electronics 736 1 0.136 1 1 31 IEEE Transactions on Power Systems 528 0 0 0 0 32 IEEE Transactions on Signal Processing 486 4 0.823 4 1 33 IEEE Transactions on Visualization and Computer Graphics 240 3 1.25 4 1.333 34 IEEE Transactions on Wireless Communications 631 0 0 0 0 35 Industrial Management & Data Systems 96 4 4.167 13 3.25 36 International Journal of Advanced Manufacturing Technology 1,420 7 0.493 7 1 37 International Journal of Numerical Methods for Heat & Fluid Flow 132 0 0 0 0 38 Journal of Lightwave Technology 691 0 0 0 0 39 Journal of Mathematical Biology 133 4 3.008 6 1.5 40 Kybernetes 101 0 0 0 0 41 Multimedia Tools & Applications 876 7 0.799 8 1.143 42 Online Information Review 68 4 5.882 14 3.5 43 Parasitology Research 532 36 6.767 56 1.556 44 Proceedings of The IEEE 155 58 37.419 91 1.569 45 Rapid Prototyping Journal 90 0 0 0 0 46 Scanning 102 0 0 0 0 47 Security and Communication Networks 475 5 1.053 9 1.8 48 Shock and Vibration 390 0 0 0 0 49 Springerplus 2,115 137 6.478 244 1.781 50 Surgical Endoscopy & Other Interventional Techniques 752 364 48.404 391 1.074
Mean 7.296 1.001 Median 0.973 1 able 7: Mention statistics for journals with Mendeley plugin S. No. Journal Name Records with DOI Found in Mendeley Coverage % Mendeley Reads Avg. Mendeley Reads
1 Acs Nano 1,285 975 75.875 61,788 63.372 2 American Journal of Sports Medicine 438 382 87.215 20,814 54.487 3 Annals of Botany 209 208 99.522 8,478 40.76 4 Bioinformatics 811 809 99.753 43,247 53.457 5 Briefings in Bioinformatics 93 93 100 9,005 96.828 6 Clinical Infectious Diseases 777 624 80.309 28,287 45.332 7 Database The Journal of Biological Databases And Curation 164 136 82.927 5,142 37.809 8 European Journal of Sport Science 155 153 98.71 8,048 52.601 9 Expert Review of Medical Devices 116 113 97.414 2,589 22.912 10 International Journal of Production Research 460 116 25.217 5,999 51.716 11 International Journal of Remote Sensing 334 84 25.15 2,037 24.25 12 Journal of Experimental Botany 514 485 94.358 24,359 50.225 13 Journal of Infectious Diseases 691 609 88.133 22,029 36.172 14 Journal of Sports Sciences 297 291 97.98 18,555 63.763 15 Journal of The American Statistical Association 161 153 95.031 3,726 24.353 16 Nano Letters 1,194 991 82.998 63,861 64.441 17 Nanotoxicology 150 93 62 3,020 32.473 18 Nucleic Acids Research 1,298 1,222 94.145 91,737 75.071 19 PLoS Computational Biology 582 580 99.656 38,243 65.936 20 PLoS Neglected Tropical Diseases 876 855 97.603 47,230 55.24 21 PLoS One 23,019 19,462 84.548 719,920 36.991 22 PLoS Pathogens 684 683 99.854 30,213 44.236
Mean 84.927 49.656 Median 94.252 50.971 able 8: Mention statistics for journals without Mendeley plugin S. No. Journal Name Records with DOI Found in Mendeley Coverage % Mendeley Reads Avg. Mendeley Reads
1 Acta Mathematica Sinica 116 6 5.172 39 6.5 2 Advances in Mathematical Physics 122 17 13.934 73 4.294 3 Aircraft Engineering and Aerospace Technology 90 5 5.556 39 7.8 4 Analytical & Bioanalytical Chemistry 837 533 63.68 11,270 21.144 5 Annals of Biomedical Engineering 305 198 64.918 7,477 37.763 6 Annals of Operations Research 312 67 21.474 2,364 35.284 7 Applied Microbiology & Biotechnology 887 519 58.512 19,765 38.083 8 Autonomous Robots 78 78 100 2,802 35.923 9 Biochemistry-Moscow 171 45 26.316 874 19.422 10 Biomed Research International 1,852 753 40.659 24,179 32.11 11 British Food Journal 188 60 31.915 2,315 38.583 12 Bulletin of Mathematical Biology 95 61 64.211 1,119 18.344 13 Canadian Respiratory Journal 95 48 50.526 1,110 23.125 14 Chinese Annals of Mathematics 73 1 1.37 2 2 15 Complexity 194 56 28.866 646 11.536 16 Disease Markers 161 67 41.615 1,520 22.687 17 Empirical Software Engineering 73 68 93.151 2,993 44.015 18 IEEE Access 815 23 2.822 5,680 246.957 19 IEEE Communications Letters 635 126 19.843 1,827 14.5 20 IEEE Journal on Selected Areas in Communications 301 109 36.213 4,445 40.78 21 IEEE Transactions on Antennas and Propagation 677 147 21.713 2,055 13.98 22 IEEE Transactions on Biomedical Engineering 269 156 57.993 7,870 50.449 23 IEEE Transactions on Communications 405 141 34.815 3,456 24.511 24 IEEE Transactions on Geoscience and Remote Sensing 573 146 25.48 3,758 25.74 25 IEEE Transactions on Image Processing 446 183 41.031 6,182 33.781 26 IEEE Transactions on Industrial Electronics 765 154 20.131 4,636 30.104 27 IEEE Transactions on Mobile Computing 224 93 41.518 2,334 25.097 28 IEEE Transactions on Neural Networks and Learning Systems 228 92 40.351 3,205 34.837 29 IEEE Transactions on Pattern Analysis and Machine Intelligence 193 142 73.575 15,652 110.225 30 IEEE Transactions on Power Electronics 736 143 19.429 5,599 39.154 31 IEEE Transactions on Power Systems 528 149 28.22 6,198 41.597 2 IEEE Transactions on Signal Processing 486 249 51.235 6,386 25.647 33 IEEE Transactions on Visualization and Computer Graphics 240 82 34.167 6,192 75.512 34 IEEE Transactions on Wireless Communications 631 207 32.805 5,641 27.251 35 Industrial Management & Data Systems 96 30 31.25 4,452 148.4 36 International Journal of Advanced Manufacturing Technology 1,420 312 21.972 7,649 24.516 37 International Journal of Numerical Methods for Heat & Fluid Flow 132 11 8.333 67 6.091 38 Journal of Lightwave Technology 691 155 22.431 3,593 23.181 39 Journal of Mathematical Biology 133 70 52.632 1,145 16.357 40 Kybernetes 101 7 6.931 137 19.571 41 Multimedia Tools & Applications 876 102 11.644 1,890 18.529 42 Online Information Review 68 28 41.176 2,031 72.536 43 Parasitology Research 532 439 82.519 8,787 20.016 44 Proceedings of The IEEE 155 92 59.355 9,812 106.652 45 Rapid Prototyping Journal 90 9 10 471 52.333 46 Scanning 102 21 20.588 338 16.095 47 Security and Communication Networks 475 66 13.895 794 12.03 48 Shock and Vibration 390 44 11.282 387 8.795 49 Springerplus 2,115 833 39.385 20,575 24.7 50 Surgical Endoscopy & Other Interventional Techniques 752 648 86.17 13,970 21.559 51 Advanced Energy Materials 419 93 22.196 7,159 76.978 52 Advanced Functional Materials 881 317 35.982 16,537 52.167 53 Advanced Materials 1,169 69 5.902 4,934 71.507 54 Applied Catalysis B Environmental 788 347 44.036 14,341 41.329 55 Biomaterials 651 449 68.971 26,121 58.176 56 Biophysical Journal 3,644 633 17.371 17,149 27.092 57 Biosensors & Bioelectronics 1,014 697 68.738 29,513 42.343 58 Cell Reports 1,100 1,079 98.091 78,027 72.314 59 Cell 661 647 97.882 168,245 260.039 60 Chinese Journal of Catalysis 239 8 3.347 204 25.5 61 Current Biology 846 821 97.045 62,638 76.295 62 Journal of Catalysis 358 87 24.302 3,852 44.276 63 Journal of Power Sources 1,651 315 19.079 14,383 45.66 64 Materials Today 96 65 67.708 7,344 112.985 65 Molecular Ecology 433 423 97.691 40,665 96.135 66 Nature Communications 3,685 3,431 93.107 300,482 87.579 7 Nature Materials 277 241 87.004 39,042 162 68 Nature 2,397 2,329 97.163 339,385 145.721 69 New Phytologist 530 523 98.679 35,671 68.205 70 NPG Asia Materials 104 98 94.231 3,446 35.163 71 Plant Biotechnology Journal 203 173 85.222 10,053 58.11 72 Plant, Cell And Environment 220 97 44.091 3,841 39.598 73 Progress in Materials Science 49 24 48.98 3,714 154.75 74 Progress in Photovoltaics 143 45 31.469 1,012 22.489 75 Progress in Polymer Science 49 22 44.898 3,287 149.409 76 Scientific Reports 21,057 14,846 70.504 604,874 40.743 77 Small 659 296 44.917 10,276 34.716 78 Statistical Methods in Medical Research 189 130 68.783 3,437 26.438
Mean36.256 37.002 Median