login26072012 12.10.12, 12:07 www.nature.com/news/rejection-improves-eventual-impact-of-manuscripts-1.11583 Odpowiedz Link czytaj wygodnie posty
kramka1 Re: Rejection improves eventual impact of manuscr 12.10.12, 12:29 Bardzo ciekawy artykuł. Szkoda tylko, ze nie mam dojścia do full text w Science, bo chetnie poczytałabym artykuł wyjsciowy... Odpowiedz Link
kardla Re: Rejection improves eventual impact of manuscr 12.10.12, 13:19 Flows of Research Manuscripts Among Scientific Journals Reveal Hidden Submission Patterns V. Calcagno1,2,3,*,†, E. Demoinet2, K. Gollner3, L. Guidi4, D. Ruths5, C. de Mazancourt3,‡ Abstract The study of science-making is a growing discipline that builds largely on online publication and citation databases, while prepublication processes remain hidden. Here, we report results from a large-scale survey of the submission process, covering 923 scientific journals from the biological sciences in years 2006–2008. Manuscript flows among journals revealed a modular submission network, with high-impact journals preferentially attracting submissions. However, about 75% of published articles were submitted first to the journal that would publish them, and high-impact journals published proportionally more articles that had been resubmitted from another journal. Submission history affected postpublication impact: Resubmissions from other journals received significantly more citations than first-intent submissions, and resubmissions between different journal communities received significantly fewer citations. Odpowiedz Link
kardla Re: Rejection improves eventual impact of manuscr 12.10.12, 13:20 With the rise of web technologies and online databases, knowledge is increasingly available regarding the process of science-making itself (1). Gathering such “metaknowledge” presents the opportunity to better understand, and optimize, the practice of research (1, 2). Citation patterns have revealed research fronts (3), maps of science (4) and citation behavior (5, 6); co-authorship patterns have revealed collaboration networks (7); and even acknowledgments can help infer contributions to science-making (8). Yet all this research rests on the emerged part of science communication: publications. These represent only the outcome of a complex process, that involves manuscript preparation, submission, peer review, and revision (9). Focusing on the final stage may give a very specific image of science (10, 11). Unfortunately, prepublication processes, which constitute a significant amount of the time allocated to research, have remained a black box for which we lack systematic data (12–15). To study prepublication history, we asked the corresponding author of virtually all research articles published between 2006 and 2008 in 16 subject categories of Biological Sciences (923 journals; table S1) whether the article was first submitted to the publishing journal and, if not, the name of the journal previously attempted (16). All evidence suggests that response bias was negligible (16). We thus retrieved the late submission history of 80,748 articles (37% of all enquiries), from which we could reconstruct the network of manuscript flows among scientific journals (Fig. 1 and fig. S2). In the network an arrow from journal A to another journal B represents a “resubmission link,” i.e., an article that was submitted to and published by journal B following submission to journal A. This network can be used to learn more about publication strategies and perceived journal importance than is available in citation networks alone (17, 18). Odpowiedz Link
kardla Re: Rejection improves eventual impact of manuscr 12.10.12, 13:20 The submission network we obtained was densely connected (Fig. 1). Most journals are thus exchanging some manuscripts with at least one other journal (16 on average; first/third quartiles: 5/21). Resubmission flows were modular in that they occurred preferentially within subgroups of journals. Partitioning of the network with a modularity-maximizing algorithm (19) revealed seven principal communities (journal clusters). These were strongly consistent with subject categories as defined by the Institute for Scientific Information (ISI) (Fig. 1 and fig. S3), confirming the expectation that manuscript resubmissions should occur mainly within disciplines. The communities were less resolved than subject categories though, and the resulting modularity value (0.5) was not extreme, indicating cross-community flow. Journals more central in the resubmission network were also those with higher impact factor (based on the ISI metric; Spearman correlation ρ = 0.55; P < 0.001; fig. S4), especially the three top multidisciplinary journals included in our study (Fig. 1). Journal importance (20), here inferred from submission patterns, is thus associated with importance as inferred from citation patterns (17, 18). This could merely result from the fact that high impact journals published more articles overall and so had more articles surveyed (fig. S4), which could increase centrality in itself. We thus studied determinants of network centrality that are not affected by the number of articles published. The first quantity is the number of times a journal was reported as an earlier choice for submission (the out-degree of a journal in the network). It increased sharply with impact factor (Fig. 2A; Spearman rank correlation; ρ = 0.55; P < 0.001), showing that high impact factor journals were more often earlier choices for submission. Furthermore, resubmission flows were highly nonreciprocal between pairs of journals (network reciprocity: 0.04) (16). This was explained to a large extent by journal impact factor: resubmission flows were oriented downstream overall, as far as impact factor is concerned (Fig. 2B). If a journal’s manuscript rejection rate increased with impact factor, a similar pattern could be created even in the absence of active preference by authors. Therefore, we built random graphs expected if resubmission behavior were independent of impact factor (16). The change in impact factor would have a very different distribution under this null scenario (Fig. 2B; dashed line), even considering the observed higher propensity of high impact journals to be the source of resubmissions (Fig. 2A). The skew in the distribution cannot be accounted for without invoking an active tendency of authors to move from high impact factor to low impact factors (Fig. 2B; solid line). Despite that, extremely negative differences (left tail of the distribution) were less frequent than expected, which indicates that authors adopt a risk-limiting strategy, and preferentially make small leaps in impact factor in the course of resubmissions. Odpowiedz Link
kardla Re: Rejection improves eventual impact of manuscr 12.10.12, 13:20 The second quantity that could contribute to journal centrality is the percentage of published articles that are first-intents (i.e., that were initially targeted at the publishing journal). Overall, 75% of all published articles were first-intents, with a range of 67 to 87% across subject categories (table S1). None of the journals we sampled was found to be purely recycling manuscripts rejected from other journals. Thus, most articles were initially targeted to the journal that would eventually publish them (this conclusion is robust to any reasonable non-response bias) (16). This indicates that authors were overall efficient at targeting their research and limiting the risk of rejection. Our results so far (Fig. 2) suggest that high-impact journals would attract many initial submissions and could select among them. In contrast, journals with low impact factors would more often receive and publish manuscripts previously rejected by higher impact journals. This would create a positive association between impact factor and the proportion of published articles that are first-intents. Surprisingly, we found an opposite pattern (Fig. 3): the proportion of first-intents decreased across the range of impact factors (Spearman rank correlation; ρ = –0.21; P < 0.001), for all except the three topmost journals in our sample (Fig. 3). Although there is a lot of scatter, qualitatively similar trends were observed within most subject categories taken individually (table S4). One explanation is that high impact journals, albeit preferred by authors, also experience stronger competition for manuscripts: they have a denser competitive neighborhood (as revealed by their more central position in the submission network; Fig. 1 and fig. S4), and together with high rejection rates, this makes it more likely to receive (and publish) resubmissions. As an example, even Nature and Science, which are certainly preferred journals in many cases, are far from publishing 100% of first-intents (Fig. 3), as each often publishes manuscripts rejected by the other. On the contrary low impact journals are more specialized, with a sparser competitive neighborhood and lower rejection rates in it: they receive proportionally fewer resubmissions from their neighbors. Odpowiedz Link
kardla Re: Rejection improves eventual impact of manuscr 12.10.12, 13:21 Impact factor was here shown to affect the submission process, but does the submission process in turn affect citation counts, and thus impact factor? Does submission history reflect the intrinsic “quality” or, in a more quantifiable way, the impact or utility of articles following publication? We compared the number of times articles were cited (as of July 2011, i.e., 3 to 6 years following publication, from ISI Web of Science) depending on their being first-intents or resubmissions. We used methods robust to the skewed distribution of citation counts, to ensure that a few highly cited articles were not driving the results (16). We controlled for year of publication, publishing journal (and thus impact factor), and their interaction (16). Resubmissions were significantly more cited than first-intents published the same year in the same journal (Fig. 4A). This is challenging to explain as the submission history of articles is not public. With most resubmissions occurring from journals with higher impact (Fig. 2A), it could be that authors are able to assess the intrinsic quality of their research and its potential impact, so that manuscripts first submitted to high impact journals, even when rejected, retained a higher propensity to be cited. However, this is unlikely as we found that resubmissions were more cited irrespective of their going up or down in impact factor (fig. S5). Several mechanisms could be involved, but perhaps the most likely explanation is that inputs from editors, reviewers and the greater amount of time spent working on resubmissions significantly improve the citation impact of the final product. There are indications of the value of peer review from publication and editorial practice (15, 21). Our results suggest it extends to citation impact. This validates the strategy of publishing groups that facilitate resubmission of declined manuscripts to other journals of the group (e.g., the Wiley Manuscript Transfer Program). Perhaps more importantly, these results should help authors endure the frustration associated with long resubmission processes and encourage them to take the challenge (14, 15). An independent effect of manuscript submission history was that resubmissions occurring between two journals from the same journal community were significantly more cited than those between two different communities (Fig. 4B). This shows that, all else being equal, changing discipline during resubmission was a risky move, yielding lower-than-average impact post publication. Hence, the boundaries self-defined by author submission behavior (the journal clusters in Fig. 1) are reasonable in that transgressing them comes at a cost in terms of impact. The network we have built from submission links (16) revealed that despite the prevalent fear of manuscript rejection (14, 22, 23), most published articles were initially targeted at the journal that would publish them and resubmission eventually paid off in terms of citation impact. Further aspects of the submission process could be investigated with this network, and it will be insightful to formally compare it to other social networks (4, 24). Adapting existing analytical methods for submission flow data could form the basis of journal impact metrics closer to author perceptions (25, 26). Surveying a broader range of disciplines to compare their standards (9) would be feasible but quite time consuming. Just as for citation analysis, a great step forward could come from concerted efforts at recording the metaknowledge contained in manuscript flows. Odpowiedz Link
kardla Re: Rejection improves eventual impact of manuscr 12.10.12, 13:26 Wyslalam PDFa :) Odpowiedz Link