The vast majority of tools used to recruit workers fail completely to predict whether someone will be successful in a job, British human resources professionals believe.
A poll by Cranfield School of Management found that a whopping 86 per cent of HR managers who take up written references do not find them useful predictors.
A further 78 per cent who use panel interviews in recruitment and the 67 per cent that use CVs do not find them indicators of future success either.
More advanced selection techniques, traditionally thought to have more validity, do not fair much better, with a high proportion of those polled having similarly negative views on competency-based interviews, assessment centres and psychometric tests.
Yet, despite all these methods being perceived as having little predictive value, organisations continue to use them as the main way to identify suitable managerial, professional and technical employees.
Of the more than 800 HR managers surveyed, 96 per cent said they used written references to some extent, 89 per cent used panel interviews, 88 per cent used CVs, 85 per cent used competency-based interviews, 60 per cent used psychometric tests and 51 per cent used assessment centres.
Dr Emma Parry, research fellow at Cranfield School of Management, said: "These results suggest that organisations are not really taking care over the choice and validation of the selection methods they use.
"They are not choosing selection methods based on how well they will actually predict future on-the job-performance," she added.
"Instead, organisations may be using particular methods because they have used them in the past or because their competitors use them. This is particularly surprising as some of these methods involve a great deal of time and expense."
The research also showed that many organisations did not verify the information that was given to them by applicants during the recruitment process.
Just 59 per cent of organisations checked qualifications from a source other than the applicant, while 78 per cent checked experience.
More striking perhaps was the fact that only 48 per cent of those organisations surveyed checked for a criminal record.
Shaun Tyson, professor of human resource management at Cranfield, said: "Given the problems of relying on one method, it is likely that employers will use a combination of selection methods to improve their selection decisions, especially for the more senior roles."
He added: "We are all aware of the disastrous consequences of inadequate background checks so I find it very surprising that so many employers are prepared to recruit people without carrying out the most basic verification of facts," he said.
Few organisations used online selection tools, with just 36 per cent accepting online submissions of CVs, 28 per cent using online application forms, 14 per cent responding to applicants automatically and 10 per cent using online tests.
More than eight out of 10 organisations provided training for interviewers or assessors 70 per cent used structured rating formats and 57 per cent used detailed competency frameworks for selection.
Most respondents saw relevant job experience as the most important factor when assessing an applicant's suitability for a job, with 70 per cent rating this as very important, added Cranfield.