"Measurement error": The number of "disruptive" patents is not decreasing at all
A year ago, a study made the rounds, according to which the innovative power of science and technology is declining. It was flawed, according to researchers.
(Bild: Dilok Klaisataporn/Shutterstock.com)
A much-cited study, according to which technical and scientific progress has produced fewer and fewer disruptive breakthroughs lately, reveals a "serious measurement error" at a crucial point. Three scientists have now determined that, at least for patents, it can be stated that research has lost none of its innovative power. On the contrary, the number of disruptive patents has actually increased in the long term, they write. Their work also proves "that the self-regulation of science works". Mistakes such as the one now discovered could happen again and again, but it is important to deal with them. The swan song for innovation received a lot of attention in January 2023.
No peak in disruption
As the group led by Jeffrey Macher from Georgetown University now recalls, the original study examined how often previous works were referred to in scientific papers and patents. If a patent was cited particularly frequently, but previous ones were not, it was considered "disruptive". However, the research group only included patents from 1976 onwards and did not consider references to earlier ones. However, if all references to patents from 1975 and earlier were ignored for patents from the early 1980s, many of the patents from that time would appear disruptive: "But not because they really are, but because many citations to the predecessor patents are not considered," Christian Rutzer from the University of Basel explains.
The team has now been able to prove that this restriction "enormously distorts" the result. As soon as older patents are included, the average disruptiveness of patents from 1980 is no longer 0.39 - as claimed a year ago - but only 0.09. By 2005, it had fallen only slightly to 0.04. At the same time, the number of highly disruptive patents had actually increased. However, the correction only applies to patents; the original research team had taken a different approach to measuring the innovative power of scientific work. The correction has now been published in the journal Research Policy.
(mho)
