Hello again from your friendly skeptical scientist, attempting to enrich and enliven people’s lives by spreading the knowledge of Science™ and contributing to everyone’s Facts To Amaze People Down The Pub list.
Today I’m going to discuss placebos and the common misconceptions people have about them and how they’re so often abused, even by rational and scientifically minded people. There’s a subtle difference between a true placebo and a simple control, and this is my attempt to explain it.
First, however, it’s important to have a proper understanding of the context here and the tool that’s been responsible for most of humankind’s recent elevation from stone-age ignorants to planet-spanning
ignorants enlightened people – the scientific method. The scientific method is a technique used to examine claims and properly determine if the claim has any truth or whether it’s an utterly unfounded boast. Its power comes from its ability to allow these investigations to be repeated by anyone and that anyone should get exactly the same outcome if the techniques are followed. It’s a simple enough process:
- Observation – Hey, when it’s cold, water goes hard
- Hypothesis – Low temperatures solidifies water
- Prediction – Lowering the temperature enough will freeze water
- Experimentation – Lower the temperature, see what happens
- Repeat – Get someone else to do the same thing and see if it matches
Falsifiability is also important – how your original hypothesis can be proven wrong. In this simplified case it would be water remaining a liquid at low temperatures. It’s important to also note that your falsifiability must be under the same conditions. Just because some numpty has put a kilo of salt in your water sample, meaning it didn’t freeze at the right temperature, doesn’t mean your hypothesis is false; the experiment wasn’t performed under the same conditions. A theory is a collection of well-tested hypothesis that come together to explain a more complex set of circumstances and not just, as it’s colloquially used, a ‘best guess’.
To properly assess the validity of your experimental results, you should include, along with the thing you’re testing, controls. Controls are almost identical to the thing you’re testing except that you already know what it’s going to do under the circumstances of the test. That way, after you’ve performed the experiment, if the control behaved as expected, you can make the assumption that the conditions were correct. As with the example above, if you wished to examine the effect of salt on the freezing point of water, you’d perhaps include water without salt in – you already know what temperature pure water freezes at. If the pure water freezes when expected, then you’re doing it right. This is a gross oversimplification, but that’s the gist of it. Establishing controls are possibily the most complex part of an experiment.
Placebos are also vital to the proper validity of many (if not most) investigations performed under the scientific method. Placebos are identical to the thing being tested except they will have no additional effect than what is expected. That last part is crucial, it’s not enough to have a blank; you need to properly assess the impact of the thing being tested and compare it against what might happen if it’s not there. Taking our water example, your pure water is a control, however your placebo would be water with a substance very similar to salt but that you know doesn’t change the freezing point (or changes it in a known way). You’re testing if salt has an effect, however you’re adding another compound to the water. If the freezing point does change, how do you know if it’s not the presence of granulated compounds making the change? You’re testing the effect of salt, not any-old-stuff.
To be effective, the person performing the analysis should not know where the placebo is, to prevent them from skewing the results (even subconsciously). To be super-dooper effective, that person should’ve received the stuff to be tested from someone else who also doesn’t know which are which. The setting up of controls and placebos should’ve been done by a third party. This complex and effective manner of assigning controls is known as the double-blind method, and is startlingly effective at destroying the psychological effect placebos can have.
A placebo is not just comparing something with something that isn’t there. Handing out a sugar pill and a sugar pill with a homeopathic remedy on isn’t a true placebo – the sugar pill is known not to have an effect already. To be a true placebo, the two sample should be homeopathic remedies, but only one has the effect you’re looking for. In drugs testing, the people receiving the placebo will usually get an older drug known to have similar physiological effects. The point is to see just how much more effective the new treatment is, not just to see if it works. Even surgery can have a placebo – in this case, all the people were operated on, but some received the new techniques and some got the old. The important thing is not to tell the patient which was which.
Placebos are horribly complicated and not always suitable for a given experiment. However, the next time someone is telling you that homeopathy/acupuncture/faith healing/(insert woo of choice here) works through the placebo effect, you know they’re talking rubbish, and even how to correct them.
For more information that’s entirely more accurate, even if it’s more wordy and boring, head off to Wikipedia. Again, as always, insult, abuse and complain in the comments below.