A number of you may have seen some articles last week in the press concerning our crossword research and, in particular, the recorded crossword solving trials we conducted in Buckingham during 2011, which some of you kindly volunteered to assist us in. I’m afraid that (the press being what it is) our research has been rather mangled, and I’d like to set out what the real position is.
On Wednesday we gave an interview to the Times, at their request, which touched upon our research in general terms. This article appeared on Thursday 27th June. We are entirely happy with the content of this report, although we very much regret the sideswipe by Tom Whipple (not us!) which was aimed at Daily Telegraph crossword solvers and which (however tongue-in-cheek) understandably offended some solvers on Big Dave’s message board.
The article appears to have spawned three bizarrely inaccurate articles in The Mail, The Telegraph, and The Express (particularly the first two). These articles appear to have been based on old data from an interview given to the University of Buckingham alumni magazine back in Jan/Feb 2012. This interview described work we were in the process of conducting at Buckingham, involving the videoing of 30 (not 750!) participants solving crosswords. It also discussed a range of potential outcomes we were hoping to explore.
Unfortunately, this conversation has been misreported as a completed and published study, and wholly inaccurate assumptions have been made about the ‘findings’ arising from it. Neither Philip nor I were consulted about these articles before publication, and all quotations are about 18 months old and quoted out of context.
Our press office has sent the following correction to the Mail and the Telegraph yesterday, with the request that it should appear prominently on the web-page of the original articles:
**********
The University of Buckingham commented:
The research reported above is an ongoing small-scale project involving 30, not 750, participants. Transcription of the video recordings is not yet complete, and no data analysis has yet been concluded of the recorded sections of the trials. Accordingly, no research papers have been published as yet in this area. The above conclusions do not reflect the views of Dr. Fine and Dr. Friedlander.
**********
We are aware that the articles have gone ‘viral’ in many countries worldwide, and that it is impossible to prevent the spread of this misinformation. But we thought it was important that you knew what had happened.
A small (genuine) paper is coming out this autumn and covers the first of the pre-solving activities at these trials. We will post information about this on the usual boards and will honour our promise to get back to all our participants individually if they requested this at the time.
We remain extremely grateful to all of you who have assisted us in our research over the years.
Kathryn Friedlander and Philip Fine
University of Buckingham
Thanks for these clarifications but I guess the only bad publicity is no publicity, as they say, so good to see crossword puzzles written about in quite a few national papers.
My impression was that the media folk wanted to emphasise the importance of practice and say how good/bad they were at solving themselves. They were not interested in a more detailed analysis of all the factors that make for successful crosswording — which I for one would have happily talked about on Radio Wales from my own extensive experience, regardless of this intriguing project. I don’t think any real harm has been done either to the cruciverbal cause or to the research project’s final results, which we all await with interest.
Is it possible to see the offending article ? Is there a link for it ?
This Telegraph article:
http://www.telegraph.co.uk/games/10145998/Help-at-last-forcrossword-solvers-who-havent-a-clue.html
says:
Are you now saying that they made that up – it’s not true – but by autumn you will tell us the real truth which is massively different but you can’t tell us right now?
Seems like a storm in a teacup to me.
Thank you, Kathryn.
As for the small item in the Daily Mail, my only thought was that it couldn’t be the whole story and it obviously wasn’t. I look forward to reading the full outcome of your research.
They would have got the 750 figure from here:
http://www.buckingham.ac.uk/latest-news/in-the-loop-issue-2/
“They have surveyed more than 750 crossword solvers and recorded more than 60 hours of people with a range of puzzle-cracking abilities.”
Who would have thought?
It looks as though the journos actually did a bit of googling instead of just reprinting your press release verbatim.
Reading the more serious blurb I would say that your sampling method is rather poor – well hopelessly skewed in fact – and certainly no good do draw any valid conclusions about crossword solvers generally – and the sample size of now 30 has to be a joke.
Don’t worry – the papers will run with your next press release. Their thirst for junk science is unquenchable.
The Daily Mail printing rubbish — well, I never.
I haven’t been following this but I don’t believe the intention of this study is to take a sample of 30 solvers at a crossword gathering and to make inferences about the population of solvers in general. Clearly, if that were the purpose of the study then there is a flaw in the sampling scheme as JollySwagman (JS) says. What I would say to JS is that a sample size of 30 might well be enough — it depends on the purpose of the study, what population you’re making inferences about and the “size” of the effect you’re wanting to detect. If you are wanting to see the different approaches that experienced solvers take to tackling puzzles compared with the average person on the street then 30 might well be enough. Similarly, if you wanted to detect differences in solving abilities between Listener solvers and Guardian solvers, for example, you would probably need a much larger sample size to pick up any signal from the noise. You can’t say “a sample size of 30 is a joke” without knowing the purpose of the study.