Bad Science

2009-11-03T14:42:00+00:00

I finally finished Bad Science last night, written by the excellent Guardian science reporter Ben Goldacre. It's a tour-de-force of rage against the combined forces of exploitative quacks, big pharma and of course the media. Throughout the book he lays down the basics of the scientific method and how it related to epidemiological research while debunking the many quack claims that get dredged up by the alternative medicine brigade. Topics covered include homoeopathy, chiropractors, faux nutritionists as well as the failings of real scientists that should have known better. It's all written in a calm measured style that doesn't come across nearly as hectoring as someone like Dawkins can do. While a lot of the basic science concepts were already familiar to me the additional details on things like selection bias and other biases less prevalent in the pure physical science experiments was concisely explained. I feel his suggestion of representing all risks in the media as natural frequencies (i.e. 1 person dies for every 100,000 horse rides*) is worthy of a campaign.

All in all I can highly recommend the book to anyone with an interest in the way medical science is reported in the media. I look forward to what he has to say in The Atheist's Guide to Christmas :-)

* I have no idea what the actual risk is, it just seemed a pertinent example given recent events

Review: Silicon Snake Oil

2009-04-09T08:23:00+01:00

I picked up a copy of Clifford Stoll's Silicon Snake Oil in a second hand book shop a few days ago. You can read the review behind the cut.

Silicon Snake Oil Silicon Snake Oil by Cliff Stoll

rating: 2 of 5 stars

I first came across Clifford Stoll while reading the excellent Cuckoo's Egg. It's a griping real life story about how he discovered and chased down one of the early Internet hackers. This is why when I was in a second hand bookstore I picked up a copy of Silicon Snake Oil. The subtitle, "Second Thoughts on the Information Highway" gives an indication about what it's about.

The first thing to note is this is a book that really shows it age. Published in 1995 it was when the Internet was moving from a cosy academic network used by scientists to the first commercial ISPs and early influx of AOLers. This when the World Wide Web was still know by the browser Mosaic. As will soon become apparent 13 years ago counts as ancient history when it comes to the 'net.

The books central thesis is one of scepticism of the promises that the advocates of the so called Information Superhighway where making. Stoll deals with the issues of information overload, signal to noise on Usenet and whether this technology will really turn people into infonauts or just passive consumers of the fire hose of information coming from another glowing box on our desks. He saves most of his reservations for the trend at the time to computerise education and worries the educational benefits of computers and 'net access are being oversold. Time and again he worries we will turn into one dimensional beings denied the "authentic" experiences of actually seeing, touching, smelling and interacting with things in the real world. There may be some interesting ideas that are still relevant for discussion today however it's hard to tell because of the numerous predictions that in hindsight completely wrong.

I don't blame Stoll for this. Predicting the future is always a tricky business. The 'net has grown up so fast and is consistently surprising the world with new inovations growing out of it. He's also not a reactionary Luddite, he "looks forward to the time when our Internet reaches every town and trailer park". However at the time he wrote this book he was clearly having a crisis of faith in what the futurists where promising.

A few illustrative predictions are worth quoting. When discussing shopping he asserts "no electronic shopping can compare with the variety, quality, and experimental richness of a visit to even the most mundane malls". This is before Amazon gave the bricks and mortar book shops a serious run for their money. He talks of the frustration of searching for information by keywords in titles of documents through various gopher services. This is before the all powerful Google "solved" the problem of search by using links to information to rank the usefulness of a page.

One thing that becomes clear is many of the obstacles he mentions has either been solved or is in the process of improving. The ease of use of computers which is another bugbear of his, usability has been late in the game of software development but people like Apple take problems like getting Grandma on the 'net very seriously. Humans have proved remarkably ingenious at solving seemingly insurmountable problems.

There are some areas he flags for concern that may still be relevant today. He wonders if the instant response of email is affecting our ability to write properly. If the ability to self publish will drown the 'net is a sea of dross. If social interactions on the screen can ever replace physically meeting people. However so much of this is mixed in with problems I know are now solved it's hard to not just write them off as excessive pessimism on Stoll's part.

In summary I would recommend reading the book if you want to remind yourself of where the 'net came from and what the early days looked like. However if your looking for a clear treatise on the potential downsides of the information world I suggest looking for a more recent book on the subject.

Alex's Adventures on the Infobahn - books

Bad Science

Review: Silicon Snake Oil