Today I rediscovered Linda Rising’s interview on “Agile 2009” conference, once dugged up by Jarek.
Punch line: third to half people on drug tests get better when given not the drug that’s tested but with just the placebo. How do we know Agile ain’t just hype / buzzwords / placebo effect itself?
The interview itself
It’s short – lasts only 10 minutes – and definitely worthwhile. I linked it in Sources, below. InfoQ has it’s very nice linked transcript feature, which is an excellent TOC for the video. For those familiar with it and really wanting just the juicy parts of even such short video: the best sections of the interview (IMO) are “Placebo Effect” and “Does it matter if it gets results?”. I don’t recommend this however. For one… come on, 10 minutes? For the other, on first few hearings where I focused mainly on those parts, my understanding of what Linda wanted to convey was seriously lacking. 🙂 I went in completely different (though in a way natural) direction.
The Placebo effect – the fascinating part
I’ll summarize how Linda learned of it, that will well explain:
- Mash series with lots of wounded, lack of morphine and “let’s pretend and give them a sugar pill”. Works for all soldiers, only one has to be ‘sugar-pilled’ twice – well, OK, “TV real”
- her friend’s case with depression being cured by placebo – so doctor’s DO prescribe sugar pills!
- actual drug trials – drugs are being verified also against “Placebo effect”. Since third to half tested people will get better just on placebo drug manufacturers wanted to know their drug is better than that.
One more example off the interview: vertebral cracks were usually cured with painful cement injection. Group of docs tried sham surgeries – quick and small incisions that did nothing. And they had same result of healing as painful cement injection to “stabilize” back that also rendered one less mobile for rest of his life.
Linda makes few remarks tying placebo-believers to Agile followers, in a positive way. Some scientist studied those susceptible to Placebo – as he called them, sheep, as she called them, believers. Turned out, sheep are more creative, open-minded and innovative (implicitly: as those going to conferences).
She admits to worrying that perhaps agile was a Placebo effect, that software development in general was just going from trend to trend, one hype after another, new buzz-words replacing old ones. To thinking of “how to measure” and making certain it isn’t so. To wanting to find out whether Agile is a sugar pill or not.
Finally, both Linda and her interviewer arrive at baffling conclusion: it doesn’t matter, as long as it works.
Since Agile is about experiment, continuously testing, series of small experiments, and since it works, it’s OK not to know.
I disagree. Obviously, if sham surgery is equally good as cement injection, then let me take sham surgery, of course without me knowing it as this’d ruin the chances of it working. That’s a given. But it obviously matters, if the standard procedure is just as good as placebo, for there may be a real cure, and we should be looking.