The Welsh Language Board has previous, as regular readers of this website know (Straight Statistics, passim). But its latest survey sets new standards.

The results appeared in various newspapers and blogs, including the BBC, the Western Mail and the Daily Post. On Click on Wales, The Institute of Welsh Affairs news analysis magazine, Meri Huws, Chair of the Welsh Language Board, rejoiced at the finding that people who don’t speak Welsh no longer see their monoglot status as a barrier to sending their children to Welsh-medium schools.  Ms Huws found this “very encouraging”, as you would, in her place.

The trouble is that the results of the survey were never meant to be published by the WLB and the press release was not authorised for public use, or so say the WLB. Pity nobody told Ms Huws.

When you see the survey details, requested from the WLB by a Straight Statistics reader, you can see why. The survey was based on just 125 respondents, parents at Welsh medium schools throughout the country, and was carried out by a PR company, MGB PR, hired by the WLB to attract attention and increase visibility for two videos presenting messages WLB wanted to spread.

In its pitch for the contract, MGB said: “Surveys and reports provide the perfect news fodder for the press” – how true! – “and are excellent at generating extensive and authoritative coverage in the media” – how half true. Extensive, maybe, but not in this case authoritative.

Question 1 asked if the respondent was a Welsh speaker. To this, 65 per cent said no. The next question asked how often respondents used Welsh in everyday life – and 72 per cent said either every day, often or occasionally. So more people use Welsh than speak it.

Question 3 asked “Does the fact that you can’t speak Welsh affect assisting in your child’s education?” To this, 66 per cent said yes. But the press release proudly asserts that 66 per cent of parents who send their children to a Welsh medium school are not Welsh speakers themselves and do not view their inability to speak the language as a hindrance to their children’s education. That’s the opposite of what the survey found,

Question 7 asked what had been the main motivation to send your child to a Welsh medium school. Location, said 5 per cent; facilities, 6 per cent, better standards of teaching 21 per cent; desire for child to be able to speak Welsh, 97 per cent. Yes, that does add up to 129 per cent. That was because respondents were able to answer yes to more than one option.

The 125 respondents gave, between them, 160 answers. The desire to speak Welsh was included by 121 of them, so it represents 76 per cent of the responses given, not 97 per cent. You might argue that all but four of the respondents included a desire for their child to learn Welsh among their reasons for choosing the school, but it would be odd to send one’s child to a Welsh medium school if Welsh-speaking were not among the features desired.

Similar quibbles apply to question 11: What do you think is the single greatest benefit of being able to speak Welsh for your child? Multiple choices were again allowed, but the percentages were calculated as if it was a single-choice question. So 61 said greater employment prospects, which is interpreted as 49 per cent. Since there were 183 responses recorded, it is actually 33 per cent. And so on.

It’s easy to see why the WLB now says that this survey should not have been published. But it was, its results were extolled by the WLB chair, it was picked up by several other media outlets, and it was only removed from WLB’s website when my correspondent raised questions about it. Congratulations to him for digging the details out, and a tiny morsel of credit to WLB for coming clean, even if it took them a month.