Navigation

Agency over-reported research publications

The Yomiuri ShimbunThe Japan Agency for Marine-Earth Science and Technology (JAMSTEC) over-reported the number of research papers it published in the last five years by about 1,100 papers, The Yomiuri Shimbun has learned.

The agency, which conducts deep-sea explorations and other research, cited mistakes made when counting the number of papers, such as counting papers with two authors twice, among the reasons for the error.

The government uses the number of papers published as an index when evaluating research institutes.

“It’s a major problem that the mistake was not uncovered,” said an official for the Education, Culture, Sports, Science and Technology Ministry, which has jurisdiction over JAMSTEC.

The agency publicizes the number of papers reviewed by outside experts in annual project reports and elsewhere. From fiscal 2011 to 2015, the agency previously claimed it had published a total of 4,062 papers, but has now revealed it published 2,948, meaning it over-reported the number by almost 40 percent.

The agency cited several reasons for the error, including double counting papers and treating magazine articles as research papers. It is possible the agency also over-reported the number of articles before fiscal 2011.

Research institutes are also evaluated based on the content of their research papers and other factors, and JAMSTEC retains a high ranking even after consideration of its corrected figures for the number of published papers.

“We take this mistake seriously and will do everything we can to prevent it from happening again,” an agency spokesperson said.

Trust in science harmed

By Sho Funakoshi / Yomiuri Shimbun Staff Writer

According to JAMSTEC, its over-reporting of the number of research papers published was due to careless mistakes.

Even if it was not intentional, an error like this could damage the public’s faith in science, because the number of papers serves as an index of a research institute’s accomplishments.

The agency employs about 1,000 people, and conducts deep-sea research and resource exploration with its research vessels, which include the manned research submersible “Shinkai 6500” and the deep-sea drilling vessel “Chikyu.”

From fiscal 2011 to 2015, the agency initially reported it had published from 667 to 988 papers per year, with the number trending downward. JAMSTEC said it discovered the mistakes when it was investigating the decline in the number of papers.

After the re-count, the annual figures ranged from 475 to 667 papers.

It is understood that the number of published papers was over-reported in each of the years.

Many research institutes, such as RIKEN and the National Institute for Materials Science, publicize these figures in reports and on their home pages.

An Education, Culture, Sports, Science and Technology ministry survey published in August found that over the past 10 years, Japan has fallen from second to fourth in the ranking of countries by number of research papers published.

This decline has became a major cause for concern and, along with other factors, focused attention on the number of papers as an important index for assessing the nation’s strength in science and technology.

“The number of published papers shows how active a researcher is. Mistakes should not be made when counting the number. This is an unbelievable error,” said National Graduate Institute for Policy Studies Prof. Terutaka Kuwahara, who is an expert in science and technology policy.

A science ministry official was also harshly critical, saying, “The organization should have a shared awareness” of the significance of the number of papers published.

To protect the public’s trust in science, there will likely be calls for strong measures to prevent such mistakes from happening again, such as establishing clear rules for counting research papers. Speech

Click to play

0:00/-:--

+ -

Generating speech. Please wait...

Become a Premium Member to use this service.

Become a Premium Member to use this service.

Offline error: please try again.