3 ms·
Necessary? No. You can do what you like. But if you want to find out what works and what doesn't, and actually get the right answer, the scientific approach is
by gjm11 14y ago
Necessary? No. You can do what you like. But if you want to find out what works and what doesn't, and actually get the right answer, the scientific approach is the only game in town.
- drcube 14y agoI'm usually the guy who defends science, but I kind of disagree with you here. Science looks for statistical trends. But I'm just one guy. Something could fail for >90% of the population and still work for me. Conversely, something could work for >90% of the population and fail for me. If the costs of trying the solution aren't high, trial and error (guided by some basic scientific theory) might be more successful than scouring the journals for whatever self help theory happens to have a little bit of clinical support these days. A lot of these theories haven't been tested, and a lot of them that have weren't too rigorous.
- freshhawk 14y agotrial and error with sample size of one (yourself) is called a bunch of anecdotes. We have plenty of evidence by now of how useful this is when looking for the truth about something. trial and error with a sample size of one is great at convincing you of something that's not true. It's horrible at everything else. The least rigorously tested theories in the softest of sciences are still orders of magnitude more likely to be right than something one person has done through trial and error experiments on themselves.
- gjm11 14y agoIf you do "trial and error (guided by some basic scientific theory)" on yourself, and do it well enough to justify believing the results you get, then you are doing scientific research. So I think we're actually in violent agreement.