Bibliometric Analysis of the Ig Nobel Prizes

Overview

The Ig Nobel Prizes are awarded each year in a ceremony at Harvard University organized by the Annals of Improbable Research. The prizes are “intended to celebrate the unusual, honor the imaginative — and spur people's interest in science, medicine, and technology.”(http://www.improbable.com/ig/) A pivotal theme of the Ig Nobel Prizes is to honor achievements that “first make people laugh, then make them think” (M. Abrahams, 2002. Ig Nobel Prizes: The Annals of Improbable Research. Orion, London). From 1991-2015 more than 250 Ig Nobel Prizes have been awarded to over 600 recipients.

The Ig Nobel prizes explicitly play to perceptions about the wackiness of science and scientists, and create much amusement in doing so. Yet, in an era where science is increasingly under pressure not just in terms of resources but also through questioning of its underlying purpose and value, the Ig Nobel Prize represents one of the ways through which the science community can find creative and entertaining ways to foster interest and education about science as well as explain how zany curiosity-driven research can actually be insightful and useful.

This project involves work that not only probes the characteristics of Ig Nobel Prize winners but also explores what can be discerned about prize winning topics and authors, their underlying “serious” research purposes, and the impacts they generate for scholarship, research sponsors and the wider public.

Participants

  • Philip Shapira (University of Manchester; Georgia Institute of Technology)
  • Jan Youtie (Georgia Institute of Technology)
  • Seokkyun Woo (Georgia Institute of Technology)
  • Yin Li (Georgia Institute of Technology)
  • Sergey Kolesnikov (Arizona State University)
  • Gennady Belyakov (University of Manchester)
  • Samira Ranaei (Lappeenranta University of Technology, Finland)
  • David Hu (Georgia Institute of Technology)
  • Marc Abrahams (Ig Nobel Prizes, Annals of Improbable Research)
  • Abdullah Gök (University of Manchester)
  • Chao Li (University of Manchester)
  • Yanchao Li (University of Manchester)
  • Fatemah Salehi (University of Manchester)
  • Milana Shapira

White Papers

  • The Ig Nobels – Who Wins What and Why? (Shapira, Gök, Chao Li, Belyakov, Salehi, Woo, Kolesnikov, Yanchao Li, Youtie).
  • Does Humor Advance Science? Evidence from Ig Nobel Prizes (Woo, Yin Li)
  • Topic analysis of Ig Nobel prize laudations (Ranaei)
  • Pride or Prejudice: How Research Organizations Respond to Recipiency of Ig Nobel prize? (Belyakov, Kolesnikov)

White Paper Summaries

The Ig Nobels – Who Wins What and Why? (Shapira, Gök, Chao Li, Belyakov, Salehi, Woo, Kolesnikov, Yanchao Li, Youtie).

The Ig Nobel Prize “honors the most eccentrically innovative minds and their unique endeavors in the sciences, arts, and humanities.” (Abrahams, 2006). Ig Nobel prizes have been awarded annually since 1991 by Nobel Prize winners at a Harvard University ceremony organized by the Annals of Improbable Research. Each year, there is an eclectic mix of laudations honoring Ig Nobel Prize winners, including for measuring brainwave patterns resulting from chewing different flavors of gum (Biology, 1997), how difficulties in recognizing one’s own incompetence leads to inflated assessments (Psychology, 2000), levitating a frog with magnets (Physics, 2000 ), showing that rats sometimes cannot tell the difference between a person speaking Japanese backwards and a person speaking Dutch backwards (Linguistics, 2007), and inventing a chemical recipe to partially un-boil an egg (Chemistry, 2015). The work that results in such prizes is typically peer-reviewed science that often only subsequently is appreciated also to be funny.

Following the development of a data set of Ig Nobel Prizes and multiple associated variables, we examine 253 Ig Nobel Prizes awarded to 595 recipients from 1992 to 2015 (we will update to 2016). We report here some initial descriptive findings. The awards are given to single individuals (for example, sole authors of papers), to multiple authors of a single paper, to two or more papers and their authors, and organizations. The most common arrangement is multi-authored papers receiving a single award, which comprised 62% of the prize recipients. Twenty percent of the recipients are involved with awards split between two papers. Sixteen percent of the recipients are single individuals. Organizations accounted for 3% of the prize recipients. In a given year, anywhere from nine to 13 prizes are awarded. The fields in which these prizes are awarded can vary but most often are for Chemistry, Medicine and Physics (25 years each). Peace awards are the next most common at 23 years, followed by Biology (21 years), Literature (21 years), and Economics (20 years). Roughly half of the years had awards for Psychology (12 years) and Public Health (10 years). Less common were awards for Mathematics (7 years), Nutrition (6 years), Engineering (5 years), and Art (3 years). Ten additional categories were offered in two years and 29 categories were uniquely offered for one year only.

Grouping these categories into broad areas, using the OECD disciplinary coding of science and technology fields, we see that 38% of the awards are in natural sciences, 20% in medical and health, 16% in social sciences, 11% in humanities, 20% to recognize “Peace” efforts, and 18% in engineering and technology. The biggest change over time in topical area is the rise of medical-related prizes in the most recent period. Most, but not all, of the prizes are for scholarly work. Seventy-four percent reference an academic paper, while the remainder refers to news articles (9%), books (7%), patents (5%), reports (3%), or other documents (e.g., artifacts, reports, theses, films, mandates, or software). Scientific papers are increasingly becoming the primary medium of this prize. In the 1991-1999 period, 60% of Ig Nobel laudations reference scientific papers. In the 2000-2007 period, 71% reference scientific papers. In the 2008 to 2015 period, 88% reference scientific papers. By region, 55 countries are represented among award recipients. Most recipients come from countries north of the equator, although there is representation in Latin America and Africa. Europe and the Americas have the largest number of recipients. These two regions account for 77% of the first authors and 73% of all authors. The US has the most recipients at 200, comprising 34% of all recipients, followed by the UK at 81 or 14% of all recipients, and Japan at 67, or 12% of all recipients. When considering the countries of the first named recipient in the award laudation, the US has the most at 32%, the UK second at 12%, and Japan third at 11%.

Does Humor Advance Science? Evidence from Ig Nobel Prizes (Woo, Yin Li)

This paper contributes to the emerging scholarship in understanding how individuals, incentives and institutions might influence the direction of scientific evolution by examining the impacts of incentives and institutions in the form of prizes on scientific evolution. The paper focuses on the Ig Nobel Prizes, because unlike most scientific awards, the Ig Nobel Prizes are awarded for non-academic merits, i.e. humorousness in research topics, independent of the awardee’s scientific achievements or influences in the field. Still, receiving Ig Nobel Prizes draws attentions from the broad scientific community, expands award winners’ reputation, and potentially gives a boost to the research area. In this regard, we conceptualize Ig Nobel Prizes as shocks to the scientific subfields where the wining scientists published. The paper uses a keywords-based method to delineate boundaries around these scientific fields rather than groupings based collaboration, co-citation, or social networks of the scientists. This keywords method relies heavily on PubMed Related Citations Algorithm (PMRA), which detects articles within the same research topical area by comparing detailed keyword information as well as relative frequencies of these keywords. Using the PMRA method, we construct a database of scientific subfields containing papers that received Ig Nobel prizes and indexed in PubMed.

We collect the prize-winning Ig Nobel data from 1991-2016. The Ig Nobel website provides detailed information about each award including winners’ names, their award laudations, awards topics, their countries of origin and affiliations, and most importantly, academic publications associated with their awards. We collect every awards information provided from Ig Nobel website, and supplement with additional information about the characteristics of the award winners. This gives us a total of 267 prize winning awards with 629 unique awards winners, where these winners range from an individual winner to a research team or to an entire organization. Of the 267 awards, 158 prize awards were associated with at least one academic publication, which is not surprising given the fact that some awards are selected purely based on their humorous nature rather than their academic contributions. From this 158 prize awards, we identify 188 unique academic publications. To delineate subfields, we restrict our sample publication to 108 publications that are indexed by the PubMed. The average number of papers within each subfield is around 90. We then match all the papers in Web of Science and retrieved their citation information.

We analyze the rate of publications, who contributes, and where the high-impact research comes from in the subfields before and after Ig Nobel Prizes. In particular, we track the publication activities of Ig Nobel award winners and their collaborators as well as non-collaborators, and we measure the relative contributions and impacts of collaborators and non-collaborators based on citations. We take advantage of the long-running and multi-disciplinary nature of the Ig Nobel Prize to show differences in dynamics across fields over time. The robustness of our results is shown through comparing to a matching sample of “boring” scientific fields i.e. fields with similar characteristics but did not receive the Ig Nobel award.

By observing the impact of Ig Nobel prizes on scientific subfields, we were able to capture micro-dynamics in scientific evolution. This result has policy implications of potential options to influence the direction of scientific fields through awards and incentives. Our results also imply that non-material incentives that provide scientists with attentions and influences such as the Ig Nobel Prizes might work just as good as material incentives.

Topic analysis of Ig Nobel prize laudations (Ranaei)

Each year since 1991, the Annals of Improbable Research gives out “Ig Nobel” Prizes in different fields for apparently trivial scientific achievements that “first makes people laugh and then to think”. The recent award for the field of psychology in 2016 was given to research titled “From Junior to Senior Pinocchio” that asked thousands of liars how often they lied and whether to believe those answers. In other words, the authors were examining the lying proficiency of people across individual’s entire lifespan. Yet, such humorous research papers impact fields of science, based on evidence such as counting the number of citations. Despite the humor in Ig Nobel prize winner papers they convey legitimate messages. Motivated to explore the characteristics of the science highlighted by the Ig Nobel prize, this study explores the content of prize winner papers. The paper uses probabilistic topic models based on machine learning methodologies that extract underlying “topics” from set of document collections to examine the extent to which there are underlying patterns in Ig Nobel prize winning research. A popular topic modeling algorithm is Latent Dirichlet Allocation (LDA), which is a generative probabilistic model, is applied here. LDA performs more efficiently in distinguishing polysemy and synonymy since it includes probabilistic models both at document and word level. LDA’s two-level analysis makes it superior to other models such as Latent Semantic Indexing (LSI) or probabilistic latent semantic indexing (PLSI). The assumption behind LDA topic models is that documents are a mixture of topics; the algorithm seeks to detect these underlying latent topics in a document collection. The topic is perceived as a distribution over a vocabulary of words.

The analysis is based on text in the laudations of 262 papers, collected from the Ig Nobel website (http://www.improbable.com/ig/winners/) from 1991 through 2016. Words are considered as a proxy that describe the emerging topics from the dataset. Ten topics out of 35 are selected for the purpose of demonstration. For instance, “topic 1 is about banana skin”, “topic 3 is about methods of trapping airplane hijackers”, “Topic 4 describes an alarm clock probably made from wasabi”, “Topic 9 appears as the relationship between dung beetles and the Milky Way!” Manual screening of the associated document for topic 9 shows the paper was about lost dung beetles that can find the right track using the Milky Way. Topic 19, 32 and 21 are more general topics about economics, life and illegal drug. Topic 17 represents a relationship between husband’s underwear and infidelity. The word fisherman is also in this topic, reflecting the semantic relationship between “husband” and “man”. This suggests that documents discussing male characters may be associated with this topic. In summary, the topics suggest a distinctive role for content dimensions of everyday life such as animal behavior, illegal/risky behavior, and life and death activities in Ig Nobel lauded research.

This research experiment has limitations. The presented result is the outcome of an experiment on a small dataset of Ignoble prize winner laudations that are very short sentences. The experiment on the small dataset shows promising, interpretable topics and removed the burden of manual assessment of 262 documents for topic detection.

Pride or Prejudice: How Research Organizations Respond to Recipiency of Ig Nobel prize? (Belyakov, Kolesnikov)

There is a long-standing controversy around research which does not seem to have an obvious practical utility, especially if it is publicly funded. Public and policymakers consider such research to be “wasteful science”, despite numerous historical accounts of “pure science” being associated with applications decades after the discovery was made. The focus of this paper is on how research organizations perceive this type of science conducted by researchers affiliated with these organizations. Some of them may recognize the potential future value while others may perceive it as a reputational threat, or even as a danger of having their public funding cut as a result, especially if accused of being “wasteful science”. A recent example of such risks is the U.S. Sen. Jeff Flake-authored 2016 report on twenty publicly-funded studies that he found “hard-to-justify;” one of these studies was the recipient of the Ig Nobel prize. Another high-profile example happened in 1995, when Sir Robert May, the Government Chief Scientific Adviser in Britain asked the Ig Nobel award committee to stop including UK researchers as awardees after public controversy around the funding sources of the work that received the prize. These controversies did not stop researchers from accepting prizes but could have affected the willingness of organizations to engage in communication about these prizes.

This work looks at how universities and research organizations decide to respond to recipiency of the prize by affiliated researchers. Do they proudly recognize it as a major achievement of their researchers and use it as an opportunity to carefully communicate the motivation and potential benefits of such research to the public? Or do they simply ignore it, hoping it will be quickly forgotten? Or do they take some action to prevent this type of research from happening under their roof. The paper argues that decisions made in relation to these questions depend on two factors: the scientific value and recognition of the work that was awarded the Ig Nobel prize, and a potential public reaction to the research as a result of the award.

To investigate the response of institutions with which Ig Nobel prize recipients are affiliated, we have collected data on scientific publications referenced on the website of the award for each prize. We operationalize scientific merit of these publications by Field-Weighted Citation Impact (FWCI), and Citation Benchmarking metric which positions the citation impact of an article against other publications of the same age and field of study. Both metrics are included in Scopus bibliometric database. The second factor - public reaction - is proxied by the number of social media mentions on Twitter, also available among metrics offered by Scopus. We limit our analysis to 62 Ig Nobel prizes awarded in 2008 and later, due to the availability of social media data dependent on the activity of Twitter user base. By adopting a two by two matrix approach, we position these publications along two dimensions according to the number of citations and Twitter mentions (or ‘virality’). We classify them into four groups: ‘clever and fun’ (highly cited/highly viral), ‘clever’ (highly cited/low viral), ‘fun’ (low cited/highly viral), ‘neither’ (low cited/low viral).

We explore the interaction of this classification with the third dimension – mentions of the Ig Nobel prize in press releases, news pages, and other communication genres on the websites of research organizations referenced as affiliations of prize recipients on the Ig Nobel prize websites. We found 130 article-affiliation pairs (because some institutions received the Ig Nobel prize more than once) for our sample of 62 articles. By using this approach, we identify how research organizations respond to this type of recognition: whether they brag about the achievement, keep a low profile/are indifferent, or employ other strategies. Overall, 56% of organizations recognized the recipiency of the prize in some form. The highest recognition (65%) is observed for “clever” articles, suggesting that it is the “safest” way for institutions to leverage the publicity gained from the award. In this case, they can easily reject potential claims of “wasteful science” by appealing to high citation impacts of the underlying publications. “Fun” and “Neither” sectors of the matrix received institutional recognition in 56% and 52% of the awards, respectively. Surprisingly, the lowest recognition - just 48% - is found for the “Fun and Clever” articles, which is comprised on only eight publications. However, by looking at the recognition patterns across organizations, we also find exogenous factors. For example, we find very few mentions of Ig Nobel prizes on websites of French institutions, which are well represented both in the full population of Ig Nobel awardees and in the “Fun and Clever” group. We also find that institutions in the United States, Canada, and the Netherlands tend to be much more open about receiving the Ig Nobel prize. Such country-level variation suggests a strong influence of institutional environment and requires further explanation.