Senior scholars stick together
Junior researchers left out of collaborations in key science fields, study finds.
6 November 2017
Ikon Images / Alamy Stock Photo
When it comes to finding a research partner, scientists gravitate towards those their age, a study of hundreds of thousands of physicists and computer scientists has found. As researchers gain publishing experience, their network of collaborators also grows, but with a seniority bias, according to the analysis by researchers at Dalian University of Technology (DUT) in the northeastern Chinese city of Dalian.
Scientists are becoming increasingly cooperative. Sociologists and bibliometricians have long been interested in their collaboration behaviour, examining it from many angles, from gender to academic discipline, scholarly interest, and institutional affiliation. Computer scientist, Xiangjie Kong, and his colleagues at DUT decided to approach the analysis of collaboration patterns from the perspective of academic age — the number of years a scholar has been publishing papers.
The role of scholars is constantly changing, as is their pattern of collaboration, says Kong. By analysing academic age, we can gain a deeper understanding of these transitions.
The researchers looked at more than 2.5 million collaboration pairings between 621,493 scientists in two public datasets: the American Physical Society (APS) and the DBLP Computer Science Bibliography. The results were published in Scientometrics in April 2017.
Kong’s team found that the majority of scientists had a short academic lifespan. They counted 33,000 physicists in the APS database who were less than four years into their career, but only 2,051 that stuck around for a decade. The dropout rate was even higher for computer scientists, with only about 100 researchers regularly publishing papers for 40 years, from an initial cohort of 60,000.
The more senior the scientist, the wider their co-authorship networks. “As academic age grows, one has both more collaborations and more collaborators in a single paper,” says Kong. Early-career scientists, however, tend to have fewer, more stable partnerships, most likely with their mentors and fellow lab members.
Overall, most of the collaborations were with scientists of a similar age to the researcher. Lu Jie, a physicist at Shanghai University attributes this to the Principal Investigator (PI) system, in which one PI often works with several similar-aged doctoral students and postdocs.
“During my postdoc, the PI was a famous scientist with many grants and academic titles,” says Lu. “I discussed and implemented most of our research with other doctoral students or postdocs. We only came to the PI when there were difficulties or the paper draft needed to be finalized.”
Yang Liying, a senior research fellow at the Chinese Academy of Sciences Documentation and Information Center, says that the study’s use of physics and computer science datasets accurately illustrates the relationship between academic age and research collaboration in these disciplines.
But, she adds, the findings do not translate to multidisciplinary fields. A researcher categorized as a new entrant to the field of physics, for example, might be a more experienced researcher in another field. “More studies are needed to examine the relationship between academic age and collaboration to find more general conclusions,” Yang told the Nature Index.