In Stephen Gorard's office at the University of Wales in Cardiff bookshelves are groaning under the weight of boxes of questionnaires.
They contain the results of interviews with more than 1,000 people in the South Wales valleys who were questioned about participation in post-school education and training.
Dr Gorard, an educational researcher, says the findings call into question some long-held myths about lifelong learning. The findings could have key relevance for the Government as it prepares to publish its White Paper on lifelong learning this February.
Dr Gorard said: "I think the idea behind this project is to at least make people step back and think again about the rhetoric of lifelong learning and the learning society."
Door-to-door interviews were held with l,l00 people aged 16 to 65. More than 30 per cent were found to have had no kind of participation in formal adult learning at all.
Dr Gorard said the results suggest that participation is determined by people's background more than by their educational experience. And he believes that just increasing opportunities for people to study and train may not be enough to get the 30 per cent back into learning.
His figure of more than 30 per cent with no kind of formal participation in formal adult learning is in line with earlier findings.
A UK-wide Gallup survey on adult participation in learning carried out for NIACE in 1996 found that 36 per cent of the adult population had had no participation in learning since leaving full time education.
But Dr Gorard's claim of participation in adult learning being determined more by peoples' backgrounds than their previous experience of education seems at odds with some of NlACE's findings.
The NIACE report on the findings, The Learning Divide, claimed that "the length of initial education continues to be the best single predictor of participation in adult learning. The more initial education and training people receive, the greater the likelihood of their learning later on".
"A particularly good or bad experience of schooling - including success or failure in examinations at 11, 16 or 18 - does not seem to make any difference to what is effectively an individual's learning pathway," he says.
"Some people stay on in full-time education after school-leaving age or move to a job with related training and others do not. It is possible to predict which is which with almost perfect accuracy simply on the basis of what is known about them when they are born."
He said the important factors are people's age, the area where they were born and their parents'qualifications.
Dr Gorard and fellow academics Gareth Rees, Ralph Fevre and John Furlong have been conducting this research - funded by the Economic and Social Research Council and by three training and enterprise councils since early 1996. The final report is due at the end of this year.
The findings are already making waves. Calls to the university's school of education last week included one from an academic due to attend a Treasury discussion on education funding, and requesting facts and figures from the project.
Stephen Gorard admitted: "I'm sceptical of how much notice is taken of research."
"It not only has to be good research, the policy-makers have to be willing to look at it, but it also has to fit in with the spirit of the age - it has to be singing a tune that people want to hear, which I think this could be."
Why choose South Wales? "What's good about industrial South Wales is the cycle of boom and bust over 50 years," said Dr Gorard. "There was the black gold era, and there was Port Talbot with the steel works, and then massive changes.
"And now we've got emerging areas like Bridgend with production, manufacturing, assembly plants - big changes. I'm sure there are other areas that produce the same, but if economic situations had an impact on people's education and training opportunities, we should be able to pick it up in South Wales.
"Also we had access to the South Wales coalfield archive in Swansea - it's a set of oral histories going back to 1890, which I think is unique. So we can put all these things together and, we hope, can tell a coherent story."
In the latest stage of the research they have begun to compile in-depth taped interviews to gauge attitudes to lifelong learning.
"My personal interest has always been with the non-participants. I enjoyed doing interviews with them perhaps more so than with people who said I did this and I did that.
"And it makes you question something else about this learning society - learning is something we do all the time. People are teaching themselves from magazines - you know, when you get the second part free and it builds into a collection. There was one guy who had plastered his house when plastering wasn't his job normally.
"His wife had wanted really ornate plaster work and he'd done it and what he'd actually done was very good. But the idea of going somewhere to learn it - maybe he wouldn't have done.
"In many cases people who don't appear on official statistics are learning a lot - but they're not participating in formal education. "
Dr Gorard believes that rather than trying to increase opportunities for everyone to learn, policy-makers should concentrate on tackling social and economic deprivation which, he says, would in turn lead to an increase in participation in learning.
"When people talk about barriers to education and training, making it cheaper, making it easier to travel, these are excellent things to do.
"I'm not criticising that - but I don't think with some of these people it would make any difference because that's not the problem. And it's not even motivation to learn. It may just be that they don't want to be involved in what's on offer.
"Sometimes what we're offering is maybe not seen as relevant to their real life. I think all of these things need to be teased out before the politicians go straight into making policy, compelling people to take part."