Volumen 1 (2016): Edición 4 (November 2016)

To address the under-reporting of research results, with emphasis on the under-reporting/distorted reporting of adverse events in the biomedical research literature.

A four-step approach is used: (1) To identify the characteristics of literature that make it adequate to support policy; (2) to show how each of these characteristics becomes degraded to make inadequate literature; (3) to identify incentives to prevent inadequate literature; and (4) to show policy implications of inadequate literature.

This review has provided reasons for, and examples of, adverse health effects of myriad substances (1) being under-reported in the premiere biomedical literature, or (2) entering this literature in distorted form. Since there is no way to gauge the extent of this under/distorted-reporting, the quality and credibility of the ‘premiere’ biomedical literature is unknown. Therefore, any types of meta-analyses or scientometric analyses of this literature will have unknown quality and credibility. The most sophisticated scientometric analysis cannot compensate for a highly flawed database.

The main limitation is in identifying examples of under-reporting. There are many incentives for under-reporting and few dis-incentives.

Almost all research publications, addressing causes of disease, treatments for disease, diagnoses for disease, scientometrics of disease and health issues, and other aspects of healthcare, build upon previous healthcare-related research published. Many researchers will not have laboratories or other capabilities to replicate or validate the published research, and depend almost completely on the integrity of this literature. If the literature is distorted, then future research can be misguided, and health policy recommendations can be ineffective or worse.

This review has examined a much wider range of technical and non-technical causes for under-reporting of adverse events in the biomedical literature than previous studies.

Based on the weak tie theory, this paper proposes a series of connection indicators of weak tie subnets and weak tie nodes to detect research topics, recognize their connections, and understand their evolution.

First, keywords are extracted from article titles and preprocessed. Second, high-frequency keywords are selected to generate weak tie co-occurrence networks. By removing the internal lines of clustered sub-topic networks, we focus on the analysis of weak tie subnets’ composition and functions and the weak tie nodes’ roles.

The research topics’ clusters and themes changed yearly; the subnets clustered with technique-related and methodology-related topics have been the core, important subnets for years; while close subnets are highly independent, research topics are generally concentrated and most topics are application-related; the roles and functions of nodes and weak ties are diversified.

The parameter values are somewhat inconsistent; the weak tie subnets and nodes are classified based on empirical observations, and the conclusions are not verified or compared to other methods.

The research is valuable for detecting important research topics as well as their roles, interrelations, and evolution trends.

To contribute to the strength of weak tie theory, the research translates weak and strong ties concepts to co-occurrence strength, and analyzes weak ties’ functions. Also, the research proposes a quantitative method to classify and measure the topics’ clusters and nodes.

To understand how authors and reviewers are accepting and embracing Open Peer Review (OPR), one of the newest innovations in the Open Science movement.

This research collected and analyzed data from the Open Access journal PeerJ over its first three years (2013–2016). Web data were scraped, cleaned, and structured using several Web tools and programs. The structured data were imported into a relational database. Data analyses were conducted using analytical tools as well as programs developed by the researchers.

PeerJ, which supports optional OPR, has a broad international representation of authors and referees. Approximately 73.89% of articles provide full review histories. Of the articles with published review histories, 17.61% had identities of all reviewers and 52.57% had at least one signed reviewer. In total, 43.23% of all reviews were signed. The observed proportions of signed reviews have been relatively stable over the period since the Journal’s inception.

This research is constrained by the availability of the peer review history data. Some peer reviews were not available when the authors opted out of publishing their review histories. The anonymity of reviewers made it impossible to give an accurate count of reviewers who contributed to the review process.

These findings shed light on the current characteristics of OPR. Given the policy that authors are encouraged to make their articles’ review history public and referees are encouraged to sign their review reports, the three years of PeerJ review data demonstrate that there is still some reluctance by authors to make their reviews public and by reviewers to identify themselves.

This is the first study to closely examine PeerJ as an example of an OPR model journal. As Open Science moves further towards open research, OPR is a final and critical component. Research in this area must identify the best policies and paths towards a transparent and open peer review process for scientific communication.

To present a method for systematically mapping diversity of publication patterns at the author level in the social sciences and humanities in terms of publication type, publication language and co-authorship.

In a follow-up to the hard partitioning clustering by Verleysen and Weeren in 2016, we now propose the complementary use of fuzzy cluster analysis, making use of a membership coefficient to study gradual differences between publication styles among authors within a scholarly discipline. The analysis of the probability density function of the membership coefficient allows to assess the distribution of publication styles within and between disciplines.

As an illustration we analyze 1,828 productive authors affiliated in Flanders, Belgium. Whereas a hard partitioning previously identified two broad publication styles, an international one vs. a domestic one, fuzzy analysis now shows gradual differences among authors. Internal diversity also varies across disciplines and can be explained by researchers’ specialization and dissemination strategies.

The dataset used is limited to one country for the years 2000–2011; a cognitive classification of authors may yield a different result from the affiliation-based classification used here.

Our method is applicable to other bibliometric and research evaluation contexts, especially for the social sciences and humanities in non-Anglophone countries.

The method proposed is a novel application of cluster analysis to the field of bibliometrics. Applied to publication patterns at the author level in the social sciences and humanities, for the first time it systematically documents intra-disciplinary diversity.