When Dr Chris Horn, chairman and CEO of IONA Technologies and chairman of the Government's Expert Group on Future Skills Needs, disclosed high third-level attrition rates last June, there were sharp intakes of breath. Until then, information on drop-out rates in the universities and ITs had been hard to come by. Horne highlighted completion rates of 80 per cent at degree level and only 65 per cent at technician level, which were impossible to ignore.

The colleges have always been coy about publishing information on attrition rates and, up until now, evidence on college drop-outs has been anecdotal. Horn's figures, though, were higher than many people expected. In particular, serious concern was expressed at the fact that up to 35 per cent of students in the IT sector were failing to complete their courses. If Ireland was to meet its present and future skills shortages, the issue had to be addressed.

In some quarters there has been a tendency to play down these figures - some observers, for example, are questioning their veracity. However, chief executive of the Higher Education Authority (HEA), John Hayden, confirms that the university retention rate is correct.

Meanwhile, Seamus Puirseil, acting director of the National Council of Education Awards (NCEA), says that, although no hard data is available, retention rates of 65 per cent in the IT sector "sound accurate."

Talk to people working in the IT sector and they'll tell you that attrition rates vary from college to college and from department to department. If this is the case, some departments and some colleges must be haemorrhaging students. Electronics and computing are among the areas which have the highest drop-out rates, sources say.

A study carried out for the Commission on the Points System identified third-level attrition rates of 26 per cent. It found that almost one-third of students with relatively poor Leaving Cert points either failed first-year exams or failed even to sit their exams. Interestingly, though, eight per cent of this group managed to pass their first-year exams but still withdrew from college. It's worth mentioning, too, that a further eight per cent of the group gained first or upper second-class honours degrees.

IF we know little about attrition rates, we know even less about the reasons why people drop out of college. Anne Carpenter, head of student services at Carlow IT, has conducted a study of first-year students who failed or who dropped out after one year. The study shows that college drop-outs tend to be younger than their colleagues, their decision to attend college is less likely to have been a purely personal one and less likely to have had parental support.

Career guidance is a vital factor - nearly two-thirds of the group received no career guidance. Almost half the students surveyed reported being less committed to their courses than their fellow students and just under one-third said they felt less well prepared for college. Over one-fifth found the transition to third level difficult.

Carpenter also found that students who drop out are more likely to have part-time jobs during term time, work longer hours and depend on income from work to finance their studies.

The Points Commission research identifies a number of different categories among drop-outs. There is the group which fails to sit exams, a second group which fails exams and withdraws from the course and a third group which passes exams and still withdraws.

Dr Kathleen Lynch, director of UCD's Centre for Equality Studies, is a co-author of the report. She points to the variations in drop-out rates from college to college and discipline to discipline.

"Drop-out rates for medicine are extremely low everywhere," she notes. "The investment to get a place at medical school is enormous in terms of time and energy." Successful completion of the course can lead to high career and financial rewards. "The reason to leave is greatly reduced," she says.

Both Carpenter and Lynch are currently engaged in a study in three ITs. "I suspect," comments Lynch, "that the study will show a complex set of reasons for students leaving."

She points to an earlier study on working-class access to higher education, which highlights the significance of guidance. "It's a major factor," Lynch asserts. "People are going to college and they haven't a clue what they're letting themselves in for. They don't realise, for example, that computer science requires a good knowledge of maths. Students make the wrong choices." Family bereavement, illness, social class and gender all play a part, she says. "People ignore the cumulative effect of all of this."

Many students coming from homes and schools where they've been spoonfed, find it difficult to handle the freedom of third-level college. Problems can be exacerbated by a range of factors including family, money and long-distance travel.

Low income students are particularly vulnerable. Part time jobs apart, they rely more heavily on college services - the library, computing and printing for example - than do their better-off counter parts, Lynch says. "If the college doesn't provide these services, it's the less well-off who suffer most."

The student unions argue that one of the biggest factors causing students to drop out is money - or rather the lack of it. "The evidence is anecdotal," admits USI education officer Darren McCarrig, "but applications to our hardship fund are increasing." These days, too, many students are lured away from completing their courses by offers of well-paid jobs.

Many experts believe that a large number of the students who leave, return to education later or move directly into lower-level courses - from degrees to diplomas for example or to PLCs.

Even with the scanty information currently available, it's clear that huge efforts must be made to improve third-level retention rates. The need for more and better guidance is particularly clear. Maybe colleges would also consider publishing attrition rates in the course prospectuses?

Anne Carpenter sees a very real need for increased student services. "Many students - particularly those in the ITs - are first generation third-level students," she says. "Their families have no prior experiences and the colleges need to take this into account and consider providing links and support services."

Traditionally, she says, support services have been viewed as add-on optional extras, but they should be regarded as integral to the academic programmes. "Students are no longer a homogenous group," she stresses. "They have different life experiences, expectations and motivations."