Ministers and exam boards have been accused of manipulating the results of the new vocational alternative to GCSE and A-levels by excluding 50,000 students from their calculations.

Gillian Shephard, the Education and Employment Secretary, trumpeted the 79 per cent pass rate for advanced general national vocational qualifications this week as “a results hat-trick” following record successes in A-levels and GCSEs.

More than 80,000 students were awarded GNVQs - double the number that passed last year.

However, researchers analysing the GNVQ data from the three vocational boards this week attacked the lack of figures on drop-outs and on students who took more than two years to complete a course.

Alan Smithers, director of the centre for education and employment studies at Brunel University, said “only by sleight of hand” could Mrs Shephard claim a 79 per cent success rate at advanced-level GNVQ. The true rate could be lower than 40 per cent, he added.

“Success rates at all three levels are based on the number of candidates tested, not on the numbers doing the courses. If you add together all the registrations and compare them with the successes, you come up short of 50,000 students.”

He also criticised the lack of interest in courses such as engineering, construction and manufacturing, which he said were vital to Britain’s continued economic recovery.

Out of 53,000 students tested at advanced level, only 188 took manufacturing, 571 engineering and 1,172 construction. Business studies continued to leap ahead with 22,851 tested, followed by leisure and tourism with 9,200.

“A major reason for the GNVQ was to have more highly trained people in the more productive areas. I see little sign of it doing so. We see a massive rise in business studies where there is already a perfectly respectable A-level.

Professor Smithers’s concern was shared by Professor Sig Prais of the National Institute of Economic and Social Research who insisted that GNVQ was inappropriate for “workplace” studies. But he was not surprised by the results which, he said, reflected a wider malaise.

One breakthrough was with one of the newest of the ten GNVQ courses - health and social care. It accounted for one in six (8,343) of all advanced GNVQ candidates tested this year and was creating a ladder to nursing and paramedical jobs, said Professor Smithers.

His criticisms have also fuelled controversy over the time needed to complete the courses. Heads of schools and colleges which run GNVQ courses insisted that the discrepancies could be explained largely by students leaving for jobs or taking longer than two years to complete the course.

John Dunford, president of the Secondary Heads Association, said: “These are not all drop-outs and failures. It is entirely appropriate for some students to take two-and-a-half, even three years to complete their GNVQ. Many leave because they find a job. This is a good thing provided they continue training. ”

But they share Professor Smithers’s concern over the lack of reliable data and accuse Mrs Shephard of putting a spin on the results in order to spread good news in the run-up to the general election.

Richard Fawcett, head of Thurston upper school in Suffolk, said: “It is very difficult to get hold of accurate figures. We need them because I am sure it will show the qualifications up in a good light.

“In my own school I have had under-achieving pupils taking a GNVQ intermediate course in one year, followed by two at A-level. They are now going to university.”

Leaders of SHA say there should be national sampling to give a broad picture of drop-out rates, time taken to complete the GNVQ and the destinations of those who leave for work or another form of study.