The joy of dodgy studies

By Brian Clegg

As a science writer I get sent lots of press releases about scientific studies. Some describe serious research, but there is always a smattering of studies sponsored by companies with entertaining PR in mind (think the ones that give the formula for making the perfect sandwich, or some such thing), rather than any scientific outcome. I’ve just received the most entertaining one I’ve had in some time, which I feel needs sharing. The press release opens with:

You might think this study was performed by a publicist with a degree in dog grooming (say), but unlike many such press releases, it does refer to an original paper. Admittedly they get that a bit wrong, saying it was ‘published on ResearchGate’ which is a portal – it was published in Current Psychology, which is a Springer Nature journal. It is also from 2022, so not exactly new, but this is because the company behind the press release, World of Card Games, has apparently conducted ‘an original analysis of the 1,000+ historical figures to identify patterns in names, birth months, and zodiac signs among geniuses’.

The ResearchGate link is a systematic review of studies that attempt to show ‘how a name influences its owner’s personality, decision-making, and life outcomes.’ I haven’t been able to obtain a full copy of the paper, but I do have a slight concern about its quality when I read in the introduction ‘To give the new coming to a survey of the forest, not the trees, we concluded the current literature to three approaches…’. It is, of course, possible that the review concluded all this stuff is timewasting nonsense, but the abstract gives no clue as to whether or not there was any critical analysis of the sources reviewed.

What’s of interest here are the key findings from the ‘original analysis’ undertaken by World of Card Games, which, according to the release, are:

• John is the most common name among geniuses, appearing 18 times (1.60%) among history’s most acclaimed inventors, scientists, mathematicians, and entrepreneurs.
• Mary leads among female names with 15 occurrences (1.33%). Other top names linked to intellectual and entrepreneurial success include Margaret (14, 1.25%), Maria (13, 1.16%), and Elizabeth (12, 1.07%).
• August is the top birth month for brilliant minds, with 117 renowned figures (10.41%) born in this month. February (101, 8.99%), November (101, 8.99%), July (96, 8.54%), and April (94, 8.36%) follow closely.
• Leo is the most common zodiac sign, with 114 notable figures (10.14%). Taurus (101, 8.99%), Aries (101, 8.99%), Aquarius (97, 8.63%), and Scorpio (96, 8.54%) round out the top five.

If we ignore star signs per se for the obvious reason that astrology is baloney, but simply consider the birth dates (that’s all star signs tell us, after all), we see relatively little variation in the lower rated months/signs – certainly too little to be statistically significant – but the tweak up for August/Leo is slightly interesting, even though it is almost certainly just the scale of outlier we might expect. It is worth noting that around this time of year (September more accurately) tends to have a birth peak, due to the celebrations 9 months earlier. I did find it odd that May doesn’t come in the top 5 months, despite Taurus (two-thirds of which is in May) being second equal – that feels dubious. 

What is absolutely hilarious, though, is the names part. Here we have to dig out that old trusty statistical mantra that correlation is not causality. Just because A (name) and B (genius) appear to be linked does not imply that A causes B. It could be pure coincidence, or equally it could be the other way around that B causes A (admittedly, unlikely here) or a third factor causes both (very likely here). In a good quality, real study, you would control for obvious ‘confounding’ factors – something else that linked A and B. 

A particularly obvious such factor jumped out to me, and with all of 2 minutes research it appears to be highly likely to be the main cause. Historically, popularity of names changed much slower than it does now. If we look at the most popular boys’ and girls’ names in the UK, for example, 100 years ago, they were John and Mary. Hmm. Do they look familiar? 

What have we discovered, then? Nothing to see here, folks. Move along.

See all Brian's online articles or subscribe to a weekly email free here

Now Appearing is the blog of science writer Brian Clegg (www.brianclegg.net), author of Inflight Science, Before the Big Bang and The God Effect.