It's I mean like uh disrespectful

It's patronising and discriminatory for academics to print respondents' inarticulacies, says Sarah Nelson

"Gorblimey guv - strike a light! I had that researcher in the back of me taxi once, and didn't 'arf give her a piece of me mind!"

When is academic research patronising and discriminatory? When you would least expect it, and when it least intends to be. Much qualitative social research among working-class and less-educated people is respectful and sympathetic. Yet the extraordinary habit of reporting taped interviews in the vernacular undoes that respect and risks creating pantomime characters.

In the past year, I've read a national study on domestic violence, a report on disadvantaged women's views of mental-health services, a study of teenagers' sexual health in a deprived area and a community development report on young people's needs: all are sympathetic, yet all reproduce respondents' local accents, poor grammar and verbal stumblings.

Thus women say: "Since he's went...ach I dinna ken, life's been that good for me and the weans." Every hesitation is left in: "Then see er I went down to the social eh and before that well my neighbour calls in and er she says..." Youngsters stumble inarticulately: "It's I mean OK like, know what I mean?" Frankly - no.

Why do researchers do this? Do they think it is somehow in touch with the working classes and demonstrative of their empathy? If they do it to remain true to what people say, why do they think it inappropriate or disrespectful to do this with other kinds of respondents? Imagine reporting interviews with old judges or rural aristocrats: "I say old man that's a bit orff, it's eppsolutely yaars since any of us did that, what?"

I wonder if researchers have considered how respondents reading their published interview quotes might feel. Imagine the offence if interviews for a study of a Chinese community reproduced people's speech problems with "l" and "r", or if the accents and phrases of elderly Asians were paraded like one of those embarrassing old television comedies.

Think of the greater shock and disgust if a stammerer were quoted as "s-s-s-s-sometimes I go to the s-s-s-surgery", or if in a study of disabled people's views of health or social services, the slurred speech of someone with cerebral palsy or the mistakes of someone with a learning disability were reproduced in the interest of realism or "truth to the facts".

In qualitative research interviews, quotes should accurately reflect people's views and feelings, and convey insight. Sometimes a dialect word or phrase will express these more powerfully than any substitute.

Occasionally, too, people's hesitations reveal something significant about them, the contradictions they struggle with (and sometimes their hypocrisy - I recall respectable Ulster politicians humming and hawing in interviews about sectarian killings). Often, the names groups of people choose to call each other (for insult or solidarity) will be important and should be reproduced exactly.

But I have yet to find that local accents, poor grammar or repeated stumbling further the interests of truth. In interviews with working-class or less-educated people, I've never seen the need to reproduce them - any more than I would parade the way a psychiatrist, social work director or headteacher has actually spoken to me.

So leave it aht, all you gritty social researchers, and give people who don't speak proper a bit of proper respect.

Sarah Nelson is research fellow in sociology, University of Edinburgh.