Observations of a lapsed scientist

I had the great pleasure of recently attending the Science Commons Symposium held at Microsoft Research. After years (decscicomlogoades really) of being away from academic science, it was refreshing to see how much had changed and how much had changed.

The one thing that really struck me is how much “trust” plays in science. The theory is that Scientists publish results, their conclusions are subject to peer review and science marches on. If only it was that clean and simple. One of the insights I came away with is that yes the result (conclusion) is important but how it was obtained (the method) as well as the assumptions inherent in the experiment are of themselves of tremendous benefit to other scientists.

To quote one of the speakers Prof. J. C. Bradley

There are no facts, only measurements embedded with assumptions.

Professor Bradley is a leader of the Open Notebook Project, whereby notebooks (including photographs and video) are put on line and thereby allow other scientists to understand the experiment not just get the results. But as Prof. Cameron Neylon stated earlier in the program.

Broadcasting is easy, sharing effectively is hard.

How to find scientific information is particularly difficult (and to be sure of its accuracy). One of the examples given was that there is a representational language in Chemistry called InChI (International Chemical Identifier). For proteins this can be extremely large, which when searched for on the internet are truncated by search engines, so a hashed version of the InChI needs to be used to search for large molecules. Trouble is it can be used to find a compound but the hash cannot be reverse engineered to the original InChI.

There was also much discussion on how the nature of scientific publishing is changing. Today a library subscription to a major journal can cost in excess of $25,000 per annum and with about 20,000 journals in circulation, it is difficult if not impossible for any library or academic institution to afford (or manage) a small subset of appropriate journals.

Other drivers include the concept that government funded research ought to be free to access and built upon and not tied up in a copyrighted paper within a journal only accessible to the chosen few who can afford it. To address this the number of online journals is rapidly increasing with more open peer review but as of yet not the prestige of the major journals.

Continue reading

Checklists and Einstein

Just finished “The Checklist Manifesto * How To Get Things Right” by Atul Gawante. check It has been a popular book on the review circuit ( see Economist, NY Times) and I think justifiably so.

I think the strength of the book comes not just from the numerous real-life evidence based examples but how the author shows how checklists are not just for simple problems anymore but can also be applied in complex situations, where the focus is not on tasks to-be-done but on communication and sign off. Turns out this social media stuff aligned with something as mundane as a checklist can be very very powerful.

It is an easy read and I would recommend it heartily to anyone with any outsider experience of complex medicine. For example, did you know the average clinic-based doctor in a single year diagnoses over 250 different conditions and prescribes over 300 different types of medication? No wonder it is so hard to prevent complications.

So where does Einstein fit in to all this. One of my favorite Einstein quotes is

It can scarcely be denied that the supreme goal of all theory is to make the irreducible basic elements as simple and as few as possible without having to surrender the adequate representation of a single datum of experience.

or as it is often paraphrased

Make things as simple as possible, but not simpler.

Turns out that his quote is one of the fundamental constructs for a good checklist.

Photo Credit:  http://www.flickr.com/photos/cayusa/ / CC BY-NC 2.0

The Scientist “Star” system

Saw a very good piece on 60 minutes last Sunday on Dr. Bob Ballard, underwater explorer, best known for his discovery of the Titanic.

If you want to read as opposed to watch, the transcripts are here (part 1 and part 2). The piece is worth watching in its entirety but a couple of things really caught my eye.

  1. While he is best known for finding the Titanic, Dr. Ballard regards his greatest triumph as the discovery of giant tube worms at 8,000 feet underwater that thrive in total darkness on hydrogen sulfide.
  2. and to the point of this entry – he was very articulate in describing the ‘star’ system in science.He also acknowledged it doesn’t hurt to be known as the guy who found the Titanic, but he said that comes with baggage. “Science is a ‘we,’ not an ‘I.’ It truly is. I didn’t do anything. We did a lot of things. But in our system, in America, we have this star-based system. Star athletes, star news people, star politicians. And stars are ‘I.’ And the academic world is really, honestly a ‘we.’”

The success of folks like Dr. Ballard , Richard Dawkins, Steve Jobs or Barak Obama is based on the endeavors of many. While acknowledging that they didn’t accomplish everything alone (or as Isaac Newton said 400 years ago – “If I have seen further it is by standing on the shoulders of giants.”) we tend to focus on one person as the smartest guy/gal in the room and totally ignore everyone else.

One wonders in this age of instant collaboration including social networks with their promise of the wisdom of the crowds what will happen to tomorrow’s stars – will they be original thinkers or the best at harnessing the thoughts of others? Who knows?

Yes Sarah, there really is global warming..

A recently published study on high and low temperatures across the continental United States has received significant attention in the US. (see Science Daily (4  stars), NY Times, Google News Search). It can best be seen graphically (copyright U.C.A.R., graphic by Mike Shibao).

Data from 1800 weather stations across the USA

While I think this graphic is compelling enough it actually understates the problem, if weather was in a relative steady state, you would expect statistically that the number of weather extreme events would tend to zero as more and more years went by. That is obviously just not happening. The Science Daily article goes into much more detail.