I know that a lot of people who read Colin’s Beauty Pages are beauty bloggers. So I thought you might like some tips on how to cut through the hype behind beauty products to get through to the reality. Here are three things to bear in mind when looking at the science stuff that sometimes gets used in the promotion of cosmetic and personal care products.
Only trials matter
Scientific theories are a big part of the business of science, but in themselves they don’t mean very much. Facts, data and evidence are the crucial things. A theory without any evidence is a like a mobile phone without any charge in its battery.
What this means is that if a product makes a claim then you should only be impressed by evidence for that claim. A good example came up recently when a number of media outlets reported on a press release about a pill that could stop your hair going grey. That sounded really interesting. I imagine nearly everybody on the planet might consider taking that one. The claim was backed up by a reference to a proper scientific journal, which is a good start. Sadly the paper itself just reported some lab work which elaborated a theory about the hair turns grey. It was a perfectly valid bit of science, but it didn’t demonstrate the claim.
If a product is marketed as doing something out of the ordinary, the only convincing reason to believe it is if a trial has been carried out in people which shows that it works.
Sample Size Is Important
Companies do carry out trials, and many of those trials are carried out to the right degree of scientific rigour. But some aren’t. Some are simply too small to show the effect claimed. Unfortunately the rules of statistics don’t allow an easy rule of thumb to be applied to tell you whether there are enough people in a trial. This is something that foxes professional scientists from time to time, and papers get published which draw conclusions that are not justified by the data reasonably frequently.
One term that gets misused both intentionally and by accident is ‘statistically significant’. All this really means is that the results are not what would be expected from chance alone. It doesn’t mean that they are meaningful to the end user. Imagine you had an anti-wrinkle cream that reduced the size of wrinkles by 2%. With enough diligence you could do a study that showed this, and a statistician would happily confirm that the result was indeed statistically significant. The end user who might have paid a lot of cash for the product would probably not regard the result as significant at all, even if they could actually see such a small effect.
If you don’t have time to go over the original report with a fine toothed comb, the best bet is just to use common sense. If they are claiming to have changed the face of the cosmetic industry but have only tried it on 20 people then they are probably being optimistic. If a well known brand has done a trial on several hundred they probably know that they need a lot of numbers to show a small effect.
Cosmetic scientists are not big contributors to the scientific literature, but most scientists spend a lot of their lives doing research that they hope is worth publishing and which contributes to scientific knowledge. As a result there are millions of papers out there full of all sorts of interesting and useful stuff.
This is generally a good thing and is one of the reasons we all live such safe, comfortable and long lives. But there is one danger, which is that with so much available it is possible to pick out facts that support a case you are trying to make. The temptation to cherry pick the bits that fit is one that people with something to sell often give in to. This is one that crops up a lot in health orientated products.
So although science is just about the best method we have come up with for learning the truth, it isn’t perfect. It can be misunderstood or manipulated. You need to keep your critical thinking skills as sharply honed when dealing with the ‘science stuff’ as you do the rest of the time.
Buy my Kindle mini-book on cosmetic ingredients