Impact What is it and how do we create it?
Joshua Knowles School of Computer Science The University of Manchester
COMP80142 — Week 2 2.19, March 19 2014 COMP80142 — Impact 2 2.19, March 19 2014 How to make an impact (case 1)
“Bouyed by the success of leaded petrol, Midgely now turned to another technological problem of the age. [...] With an instinct for the regrettable that was almost uncanny, he invented chlorofluorocarbons, or CFCs.”
Bill Bryson, A Short History of Nearly Everything
COMP80142 — Impact 3 2.19, March 19 2014 Lecture Summary
What is impact?
How to have it, including some well-known cases
Impact in EPSRC and REF terms (brief)
Against impact
COMP80142 — Impact 4 2.19, March 19 2014 What is impact?
[ Discussion (5 mins) ] Make a list of measurable or concrete ways in which your work (specifically research papers) could have impact.
COMP80142 — Impact 5 2.19, March 19 2014 What is impact?
You get a patent. Careful: you shouldn’t publish a paper first
Your idea is implemented in a small company
Your idea becomes part of the standard C library, or is behind all internet security globally
A technique many people use is due to you, even named after you
You are on QI, Stargazing Live and used to be in a rock band
You teach, inspire, and enthuse the general public via popular writing or other media
COMP80142 — Impact 6 2.19, March 19 2014 Your review or survey is used by everyone in the introduction of papers
Your paper gets many citations. Is that impact?
You win a prize. Is that impact?
Your paper is published in a prestigious journal. Is that impact?
You are made professor of computer science at Stanford. Is that impact?
COMP80142 — Impact 7 2.19, March 19 2014 What is impact?
EPSRC carves it up like this
COMP80142 — Impact 8 2.19, March 19 2014 How to make an impact (case 2)
Invent The World Wide Web Tim Berners-Lee may not have fully appreciated the impact that the invention of the World Wide Web would have, but he did have an idea; he was driven by doing something that would help scientists at CERN (and more widely) share and organize data more readily.
He was tool driven.
The impact was from this . . .
COMP80142 — Impact 9 2.19, March 19 2014 . . . to this
COMP80142 — Impact 10 2.19, March 19 2014 How to make an impact: the hard route
Do outstanding science, stupid.
But that may be a hard trick to pull off, especially if you miss/missed any of Research Methods I, II or III.
Good advice, however, can be found here:
Advice to a Young Scientist (Peter Medawar) The Act of Creation (Arthur Koestler) Letters to a Young Scientist (Edward O Wilson)
COMP80142 — Impact 11 2.19, March 19 2014 How to make an impact: the easier route
Write well
Use rhetoric to persuade
Be generous to others in your field
Be bold
Tackle problems of wide interest
Create tools or data
Explain ways that your idea could be developed further.
COMP80142 — Impact 12 2.19, March 19 2014 How to make an impact: the easier route
Write well
Use rhetoric to persuade
Be generous to others in your field
Be bold
Tackle problems of wide interest
Create tools or data
Explain ways that your idea could be developed further.
COMP80142 — Impact 13 2.19, March 19 2014 Be generous to others in your field
The Young Upstart Scenario
You are new to the field. You write a paper that states that some big- name professor is wrong about everything he’s published in his long and illustrious career. What is the chance you get it published? If you get it published, what is the chance you make friends? What is the chance you are highly cited?
Don’t you think it would be wise to take a softly, softly approach?
COMP80142 — Impact 14 2.19, March 19 2014 More generally,
Don’t say: There have been some limited and isolated uses of hypertext (Berners-Lee, 89), but we here invent a groundbreaking revolution that eclipses previous achievements by allowing hyperlinks to appear pink when clicked.
Do say: “I have stood on the shoulders of giants” [although probably not so dramatically]
Do this because it is the right thing to do. Respect others’ work as much as you can.
COMP80142 — Impact 15 2.19, March 19 2014 Be bold
At the same time as being generous and humble, you do need to be bold.
This means your paper should come down clearly on one side or the other of some question, challenge or technical mystery.
COMP80142 — Impact 16 2.19, March 19 2014 Imagine after your paper is published. Will someone else be able to write of you (Jones):
Jones presented evidence [or a well-argued case] in support of X. Peters [some other person] took the opposite view, arguing Z.
Or will they have to write:
Jones thought that the question as to whether X, or Y or Z is true is debatable.
Most people would find the latter weak writing about weak science, and will cut Jones out of the story.
COMP80142 — Impact 17 2.19, March 19 2014 Create tools or data
Papers that provide tools (software), data (benchmark problems or results) or other useful resources to the community are cited more, as a group, than those that don’t.
Example (from my field):
Of all JE Beasley’s (∼ 100) papers, his most cited, and the one for which he is perhaps best known describes the OR-Library, a resource to distribute test optimization problems by email.
COMP80142 — Impact 18 2.19, March 19 2014 Discussion point
[5 minutes] Should science be pursued for its own sake, or only for its impact?
Subsidiary questions:
Is impact predictable or even measurable?
How could you quantify the impact of Brian Cox?
COMP80142 — Impact 19 2.19, March 19 2014 How to make an impact (case 3)
No pun intended:
Although obviously curious, Leonardo was a paid inventor for much of his life.
COMP80142 — Impact 20 2.19, March 19 2014 Impact for the EPSRC
Proposals for grants or fellowships are required (by EPSRC at least) to have an Impact Summary and a Pathways to Impact statement.
More information on these can be found here http://www.epsrc.ac.uk/funding/guidance/preparing/Pages/economicimpact.aspx
COMP80142 — Impact 21 2.19, March 19 2014 Impact and the Research Excellence Framework (REF)
The research councils now attempt to measure research impact for each unit of assessment (e.g. Computer Science) within each University. Funding is based partly on this score.
The REF definitions explain that impact is change or benefit to the economy, society, culture, public policy, . . .
So publishing in “high-impact” journals is not enough, and nor is being highly cited.
See more details here http://www.thinkwrite.biz/pdfs/quick impact.pdf
COMP80142 — Impact 22 2.19, March 19 2014 An extract from an ‘Impact Case’ for REF can be seen here http://www.ref.ac.uk/media/ref/content/background/impact/physics.pdf
COMP80142 — Impact 23 2.19, March 19 2014 COMP80142 — Impact 24 2.19, March 19 2014 Against impact
Changes to EPSRC funding policy over the past two years [. . . ] are rapidly undermining the councils ability to fund the type of curiosity-driven research that won Geim and his colleague Konstantin Novoselov the 2010 Nobel Prize for Physics. These changes are driven to a large extent by the fact that all grant applicants are required to submit a two-page Pathways to impact statement.
Philip Moriarty University of Nottingham, UK
COMP80142 — Impact 25 2.19, March 19 2014 Against impact
Several scientists have stated that science should be driven by curiosity.
Examples of discoveries that resulted primarily from curiosity are the following:-
Special and General Relativity
Penicillin
The Higgs Boson
Green Fluorescent Protein
Wireless LAN (resulting from radio astronomy)
COMP80142 — Impact 26 2.19, March 19 2014 Against impact
The Higgs Boson — small gravity, big impact
“Today, I wouldn’t get an academic job” (Peter Higgs)
‘Higgs said he became “an embarrassment to the department when they did research assessment exercises”. A message would go around the department saying: “Please give a list of your recent publications.” Higgs said: “I would send back a statement: ‘None.’ ” ’
http://www.theguardian.com/science/2013/dec/06/peter-higgs-boson-academic-system
COMP80142 — Impact 27 2.19, March 19 2014 Summary
Impact is a buzz word in science and academia
We are supposed to pursue it and predict it, yet it is clear the best science is done out of curiosity or for its own sake in many cases
However, thinking about the impact of your work is still worthwhile as long as you don’t lose your curiosity, and don’t spend all your time checking whether your ideas made someone else wealthy or famous
Impact is not new. Science has always needed funding, and funders tend to want results
COMP80142 — Impact 28 2.19, March 19 2014