This is an announcement of a commercial reporting service, not the results of a clinical study or controlled evaluation of research impact. According to the release, the reports are designed to give medical affairs teams a more detailed view of how their research output and independent medical education activities are received. The service will draw on Dimensions and Altmetric data and present the results through narrative reports.

The companies say this information can help teams understand the impact of their work and feed into publication strategy. The release does not provide a completed example report, performance measurements from the new service, or measured changes associated with the partnership. It also does not specify the datasets, search parameters, reporting periods, sentiment method, benchmarks or procedures that Sunmedia will use in each customized analysis.

Sunmedia serves around 2,000 corporate customers across several sectors, with pharmaceutical companies forming its core market. The release says that customer base includes most major pharmaceutical companies in Japan. Established in 1964 and headquartered in Tokyo, Sunmedia provides research-content tools, electronic-journal hosting, rights permissions and single-article sales to university libraries, hospitals, corporations and government agencies across Japan.

It also connects scholarly publishers, artificial-intelligence companies and research-solution providers with the Japanese scientific community. These activities place Sunmedia within Japan’s scholarly-information and corporate-content market, which is the local market addressed by the partnership. Koji Yamashita, chief executive of Sunmedia, said medical affairs teams in Japan had long needed a clearer way to demonstrate the value of their work to leadership.

He said combining Dimensions and Altmetric data with a narrative-driven approach gives the companies a way to help teams tell that story with evidence behind it, adding that this type of reporting had not previously existed in the market. Gareth Williams, director of partnerships at Digital Science, said the partnership would provide medical affairs teams with insight into how their research output is received. He said it would help teams justify their impact internally and measure it externally in clear, accessible terms.

Digital Science describes itself as an artificial-intelligence-focused technology company serving researchers, universities, governments, funders, industry and publishers. Its portfolio includes Altmetric, Dimensions, Figshare, IFI CLAIMS Patent Services, metaphacts, Overleaf, ReadCube, Symplectic and Writefull. The announced collaboration applies these research-information resources to reporting for Japanese medical affairs teams through Sunmedia’s local customer operations.

Although clinical guidelines are cited by the companies as an example of a real-world outcome that can be connected to research activity, the release does not report guideline findings, establish causal links between publications and guidelines, or show that the service changes clinical practice or health outcomes.

Subject of Research: A commercial partnership for reporting research impact among Japanese medical affairs teams

Article Title: Digital Science and Sunmedia partner to bring research impact reporting to Japan's medical affairs teams

Article References: Digital Science and Sunmedia partner to bring research impact reporting to Japan's medical affairs teams. Original publication (Bibliographic details unavailable.)

Image Credits: AI Generated

DOI: Not provided

Keywords: Digital Science, Sunmedia, medical affairs, research impact, Dimensions, Altmetric, Japan, clinical guidelines, publication strategy, independent medical education, Digital, Science

The research, led by Pietro Pollo and Michael Kasumovic of the University of New South Wales, represents the most comprehensive assessment of taxonomic bias ever conducted in the field. Previous studies relied on small samples of journals and manual classification of papers, leaving open the question of whether the patterns documented in the early 2000s had persisted. To overcome these limitations, the team analyzed 1,383,803 papers from 612 journals, spanning literature published as far back as 1877, and tracked how the representation of different animal groups changed between 2001 and 2025.

Classifying papers by hand at this scale would have been impossible, so the researchers built a keyword-based algorithm trained on nearly 5,000 manually classified papers. The algorithm drew on the GBIF Backbone Taxonomy and a customized vocabulary of taxonomic terms to identify the animal focus of each paper from its title, abstract, and keywords. When tested against 1,815 papers the researchers had classified independently, the algorithm achieved at least 97.2 percent accuracy for every animal group, with an F1-score of 91.2 percent. The team also verified that the algorithm rarely produced impossible labels, such as identifying a paper in a mammalogy journal as being about something other than mammals.

The scale of the bias is striking. Of the 991,894 papers involving animals, 46.1 percent focused exclusively on invertebrates and 45.1 percent exclusively on vertebrates, figures that appear roughly balanced until one considers that invertebrates account for approximately 95 percent of animal species. When the researchers compared the observed proportion of papers on each group with what would be expected if research effort were distributed in proportion to species richness, insects emerged as the most underrepresented group relative to their diversity, while mammals and fish were the most overrepresented.

The study also documented how bias is embedded in the structure of the field itself. Nearly 40 percent of the 612 journals analyzed had explicit taxonomic restrictions, and a striking 42 percent of these specialized journals focused on arthropods. The researchers found that species-rich invertebrate groups such as insects and arachnids are often relegated to these specialized outlets, which likely reach narrower audiences than general ecology and evolution journals. This means the true extent of vertebrate favoritism may be even more pronounced from the perspective of an average researcher scanning the broader literature.

Perhaps most disappointingly, the overall pattern has shown no improvement since taxonomic bias was first formally discussed in 2002. The proportion of papers on vertebrates across the literature as a whole remained stable over the 25-year study period, and in some subfields it actually increased. In aquatic journals, for example, fish rose from 35.8 percent of papers in 2001 to 49.7 percent in 2025. In behaviour, conservation, and general ecology and evolution journals, mammals and birds remained particularly dominant. The only subfield where invertebrate research increased was zoology journals, likely because these outlets publish much of the work describing new species, the vast majority of which are invertebrates.

The researchers also examined whether journal prestige metrics were linked to taxonomic focus. Papers published between 2021 and 2025 in journals with higher impact factors and citation indicators were more likely to focus on vertebrates, although the relationship was not uniform across subfields. Interestingly, birds were equally likely to appear in high-impact and lower-impact journals, suggesting that the prestige advantage accrues to vertebrates as a broad category rather than to any single charismatic group.

The attention gap proved even more consequential than the publication gap. Compared with papers on invertebrates in the same journal, papers on vertebrates received on average 10.2 percent more citations, 43.1 percent more mentions in online documents such as news outlets and social media, 24.2 percent more readers on the reference manager Mendeley, and 45.1 percent higher Altmetric scores, a weighted measure of overall online attention. These findings align with previous research showing that vertebrates attract more attention on platforms like Instagram and Wikipedia, and even on government-curated imagery such as postage stamps. Within the attention metrics, birds often attracted the greatest online buzz but received the fewest citations, revealing that public appeal and academic influence do not always coincide.

The authors caution that their within-journal attention analyses are correlational and cannot fully rule out confounding factors such as paper topic. Still, they argue that the patterns likely reflect deeper societal biases, since surveys consistently show that people value vertebrates more than invertebrates, a perception that editors, reviewers, and funding committees may unconsciously apply when judging research appeal. The consequences extend beyond academic vanity: taxonomic bias can stifle medical and engineering innovation by limiting biomimetic inspiration from understudied organisms, and it undermines the theoretical generality that ecology and evolution depend upon.

To break the cycle, the researchers propose concrete interventions within academia. They urge journals and funding bodies to systematically collect and publicly report the taxa studied in submitted manuscripts and funded projects, to incorporate taxonomic diversity into guidelines for awards, hiring, and publication decisions, and to establish targeted financial and publishing incentives for research on neglected organisms. They also emphasize the need to revive descriptive disciplines such as taxonomy and natural history, since hypotheses cannot be tested about organisms that have never been described. While such measures may seem ambitious, the authors note that similar skepticism once greeted diversity initiatives that are now widely accepted, and they hope their findings will prompt institutions to extend their commitments to equity to the tree of life itself.

Subject of Research: Taxonomic bias in ecology and evolution research literature

Article Title: Taxonomic Bias Persists in Ecology and Evolution Research Despite Decades of Concern

Article References: Pollo, P., & Kasumovic, M. M. (2026). Taxonomic Bias Persists in Ecology and Evolution Research Despite Decades of Concern. Ecology and Evolution, 16(10), Article e74450. https://doi.org/10.1002/ece3.74450

Image Credits: AI Generated

DOI: 10.1002/ece3.74450

Keywords: taxonomic bias, ecology, evolutionary biology, vertebrates, invertebrates, journal impact factor, Altmetric, biodiversity, conservation, entomology, scientific publishing, citation bias