25 May 2022

Nobody is perfect and errors do, inevitably, creep in to even the best publications. In the February issue of Physics World, David Marshall was right to highlight this problem in the context of school textbooks, but these mistakes are as nothing compared with the persistent misinformation that floods the Internet. The problem is particularly acute at the intersection between science and popular culture.

A few months back, my year 8 students (13–14 year olds) were trying to get to grips with energy in food. Teenagers are far more familiar with calories than joules in the context of food, but they were thrown into confusion by the bombshell that what people often call "calories" are actually kilocalories. "So does that mean somebody who eats a thousand calories is really eating a million calories?" a student asked. "That’s so wrong," another countered. "If you ate a million calories you would be huge."

To explore this issue, we turned to the Internet. Our first search told us that an apple contains 14 calories. I said that’s way too low and the class told me my mistake was to use Bing (the school’s default search engine): apparently I have to look on Google. Here the top hits revealed that an apple typically contains around 50 calories. The students were much happier with the higher number, but I pointed out that’s hardly any better and the right answer is probably 50 kcal. As it happens, the NHS website gives 47 kcal (196 kJ).

My heart sank when another student asked, "Aren’t calories and kilocalories the same thing?" Thankfully, a few of them were starting to correct this error among themselves, and I reiterated that the two units – as written on the side of the cereal box I was showing them – are not the same thing and that the problem is with our everyday language.

"The whole Internet can’t be wrong," insisted one pupil, trying to resolve what had become a very animated debate. Then another student joined in to ask, "What about a McDonald’s chicken mayo?" We turned again to the great god Google and the first hit told us that this favoured treat contains 413 calories. I scrolled down because a link to McDonald’s own website had caught my eye, where a mayo chicken is declared to contain 319 kcal. "So McDonald’s has got it right," somebody observed, to my great relief. But with so much inconsistency and inaccuracy online, is it any wonder that students struggle to separate true facts from accepted norms?

"

Try searching for diagrams showing "the dispersion and recombination of white light through a pair of prisms". Sadly, the vast majority of images returned are fundamentally flawed

This was a particularly productive classroom debate, and there would be a good case for saying that a lot of learning took place in that lesson, but try searching for diagrams showing "the dispersion and recombination of white light through a pair of prisms". That’s pretty specific, so you would expect to get a suitable calibre of hits showing a symmetrical arrangement of prisms and spectra, separated by a biconvex lens, as sketched by Isaac Newton in part II of his First Book of Opticks, Proposition XI. Sadly, the vast majority of images returned are fundamentally flawed.

A lot of the diagrams show the two prisms pointing in opposite directions, relatively close to each other with no lens in between. While this set-up can appear to recombine the light, close inspection would reveal that the "recombined" light is actually a rainbow. In some of the images, the spectrum between the two prisms has beams of parallel colours, which at least provides the symmetry expected from the reversibility of ray diagrams in geometric optics. Other diagrams correctly show the emerging spectrum diverging from the first prism, as is the common result when searching for the effect of a single prism, but then have some colours magically refracting beyond the normal to achieve recombination inside the second prism. Contrary to my student’s suggestion, the Internet clearly can be wrong – and overwhelmingly so.

One answer would be to direct students towards reliable websites, such as BBC Bitesize and NASA, but this is akin to running a library based on books from only two publishers, which would be unhealthy. Encouraging students to look at the domain name of a website before accepting its information as accurate is a good step forward but this requires a certain level of web literacy (which most have) and a willingness to look past the first few results (which is rarer).

Adding inurl:ac.uk can be a very useful filter for A-level students who are seeking high-quality sources but this isn’t often helpful for those who are just starting out on their voyages of scientific discovery. Dedicated search engines and portals that return hits from reliable and accessible sources are very thin on the ground (refseek.com being one) but they are not Google so, no matter how good such services may be, few students will remember to use them.

Of course, there are many cases where learning has to be simplified to make it accessible. Nevertheless, there is a difference between simplification and persistent errors. I am curious to know how prevalent these problems are and I would therefore be interested to hear if other readers have their own examples of common online misinformation.

As for the numerous incorrect representations of the dispersion and recombination of white light, maybe they all stem from Pink Floyd’s cover for Dark Side of the Moon and somebody deciding to flip a mirror image? Whatever the reason, Isaac Newton would surely be turning in his grave, if only he had access to the Internet.

Jon Tarrant is, most recently, a former head of sciences who has taken early retirement to pursue freelance work, e-mail jontarrant@cantab.net