Daily Archives: August 21, 2008

Practical, but not Practical

Posted on by

The practical joke arms race of Caltech, MIT and … UBC? Extreme Engineering

In this arms race, UBC is the third superpower. One of its most sophisticated feats also took place on the Lions Gate Bridge, in 1988. Electrical engineer Johan Thornton, now a contract engineer in his late 30s, decided that he wanted to make the bridge lights —all of them—blink. Thornton will only broadly describe the hack, but he hints that the low current of the bridge’s daylight sensor was crucial. For hours, people assumed that the blinking bridge lights were broken. Then the crew of a passing cargo ship reported that the pattern was Morse code: “UBC engineers do it again.”

I had no idea. I spent two and a half years at TRIUMF, on the edge of the UBC campus, but don’t recall hearing about any adventures in my time there.

Good Argument, Bad Argument

Posted on by

Ran across the tube containing Standards in Science Blogging and My Inbox. I’m interested in standards of science blogging, so I gave it a read. The author almost gets it right when talking about the right way and wrong way to support your argument.

There is a right way and a wrong way to buttress one’s viewpoints on controversial issues involving science and society.

The right way is to do a comprehensive search of the literature on the topic and to find a group of peer-reviewed articles that support one’s argument. In a popular article, it’s OK to also quote popular sources, but if the subject is science, the focus should be on peer-reviewed mayerial.

I think you need to take it one step further. Finding articles that support your argument is the lawyer’s way of making a case. The scientist looks at all of the material, or at least a reasonable sampling of it. In any widely-researched area there will invariably be some literature that is unsupportive, contradictory or at least ambiguous, and it is not scientific to cherry-pick results. This is just the nature of, well, nature — you get statistical results, and sometimes those results are the outliers rather than the average.

So make sure it’s carefully-done research (peer-review helps with that). But survey the whole body of it, and make sure the science really is supporting you. There are people who will point to a poorly-done study and build a position from it, oblivious to the fact that it contradicts mounds of other works — these are not good arguments.

I think it that science bloggers and journalists should work toward a standard of ethics that their scientifically-related posts and articles will contain at least a minimal number of links or citations to peer-reviewed material.

Obviously, if blog posts aren’t about science, there’s no such need for literature citation.

I think this is true, remembering the context of discussions of science & society. One also needs to remember the difference between fact and opinion. There are quite a few people out there who post their opinions as if they were facts, or dismiss facts as if they were opinions.