This forum is about wrong numbers in science, politics and the media. It respects good science and good English.
I noticed this article in the Daily Telegraph a few days ago. The Telegraph seem to be running some sort of campaign against Donald Trump. In this case they've noticed that Trump has a rather strange signature and they decided to call in a "graphology expert" to analyse his character and also look at the signatures of various other politicians.
The problem with this is that graphology doesn't really work. A lot of people in the UK mainstream media seem to think it works, probably because they've heard of firms, particularly outside the UK, supposedly employing graphologists in the activity of recruiting people.
James Randi investigated whether graphology works, and demonstrated on a TV programme (which I think was made in the UK and broadcast in the early 1990s) that it's not very effective. After searching around a bit, I managed to find Randi's investigation of graphology in a Youtube video:
The graphologist, Duncan McIntosh, presumably one of the UK's top graphologists of the time, was asked to identify what jobs five women worked in from samples of their handwriting, and he only got one out of the five right, no better than random guesswork. (I suppose McIntosh would claim that most of the women were working in the wrong jobs)
Randi mentions a statistic near the start of the clip where 85% of European firms supposedly use graphologists in the recruitment process. If this statistic is correct, it sounds like another good reason to get out of the EU.
I thought I would do a bit of checking on the "85% of European firms use graphology" statistic mentioned in the previous post. Randi quotes it in the YouTube clip, and I seem to remember the same figure being quoted in every newspaper article you might read about about graphology in the 1980s. It sounds implausible, a bit like a "10% of the population are gay" type statistic.
Somebody from the University of Westminster (which I didn't realise had a university) seems to have come up with better researched statistics for the use of graphology in various countries:
It looks like the use of graphology is most common in France and Switzerland, and it could be used more often than not in those countries. It also seems to be well accepted in Belgium. Outside those countries it is not that well accepted. In the UK, usage could be as high as 6% of medium and large sized firms.
For amusement I thought I'd also have a look around for any statistics claimed by UK graphologists (who will obviously have an interest in exaggerating the figures to create more work for themselves), and found this article:
In the article a graphologist called Erik Rees claims that up to 30% of UK firms use graphology in the recruitment process. Rees also claims that the SAS use graphology.