Are objective measures the answer to a-level grade inflation?

As the annual debate over A-level standards gathers momentum, objective measures of work applicants show verbal reasoning skills amongst graduates have risen for the first time since 1999 but numerical skills continue to decline.

According to data from nearly 25,000 test takers sitting tests from SHL Group plc, the world leader in objective assessment, verbal skills of candidates in 2004 were marginally better than those of candidates in 2003. However, numerical skills continued a six-year-long trend of decline. In 1996, candidates were answering an average of 50 per cent of numerical questions correctly but in 2004 this had fallen to an average of 38 per cent correctly answered.

The following graph shows the relative decline of verbal and numerical performance of candidates over the past 8 years. 1996 average scores for both numerical and verbal tests have been re-based to 100 for ease of comparison.

The decline in verbal reasoning is less pronounced than that of numerical, but it still represents a fall from an average of 65 per cent of questions answered correctly in 1996 to just 56 per cent answered correctly in 2004.

Kevin Kerrigan, managing director for SHL UK commented: The annual debate on A-level standards is really a moot point for most employers. With more and more job applicants having virtually identical academic records, including high-grade A-levels and 2.1 degree passes they are finding it increasingly difficult to differentiate between applicants based on academic achievement. Our tests provide objective assessments of candidates, not only rating them against their peers, but also allowing like-for like comparisons with previous years’ intakes.

The SHL tests, designed to identify how well an applicant can reason from written and numerical information, provide the candidate with passages of text or tables, graphs and charts of numerical data from which they must use to select the correct multiple-choice answer.

These assessments, commonly used by many of the UK’s leading organisations, use carefully constructed sets of questions, and are timed so that only exceptional candidates will have time to attempt all questions. SHL’s data derives from the average number of questions answered correctly by test takers in a given year thus providing a good indication of average ability levels in these core areas.

The tests are not specific measures of literacy and numeracy but they do involve comprehension of written information as well as some capability with basic arithmetic calculations such as averages, ranges and proportions. The verbal reasoning requires a reading level equivalent to Year 12 or ’sixth-form’, whilst the numerical reasoning requires a standard of mathematical knowledge roughly equivalent to GCSE.

SHL, a pioneer in the field of occupational psychometrics has ensured that the standard and difficulty of these tests has remained constant over the last decade allowing employers to make like-for-like comparisons and to assess current graduates against the scores of previous hires. Unlike A-levels they are scored on a normative basis allowing organisations to differentiate between the truly exceptional and the merely proficient.