In the Ospre Part I examination, whether it is to the rank of Sergeant or to Inspector, candidates will have to answer 150 multiple-choice questions. However, only 140 of those questions will count towards a final mark. This is because 10 of the questions in the examination are validation questions.

Validation questions are 'test' questions inserted at random points within the examination paper. Effectively, they are included in an examination as a test of the question itself (and not of the candidates' law knowledge) to establish if the actual question is good enough to be included in a future examination. Another way of thinking about these questions is that they represent a form of quality assurance for questions. If, after validation, the question matches set statistical criteria then it will be introduced into a future examination as a 'real' question and will count as one of the 140 questions that make up a candidates' final score.

Many candidates criticise the inclusion of validation questions and, on the face of it, it is easy to understand why as these 'test' questions can have a detrimental effect on candidate performance. For example:

Considering this scenario, it is reasonable that candidates may strongly feel that verifications questions should not be placed in an examination. However, there is a far weightier counter-argument in favour of the inclusion of verification questions in the examination.

How would a candidate feel if they revised over a period of six months, spent a considerable amount of time, effort and money in the process and then took a 140 question examination where the questions had just been thrown together because they were relevant to the syllabus but had never been strictly analysed to make sure they were fair and then, as a result, that candidate failed the examination? Can you imagine the amount of thoroughly justified and warranted complaints this would create? This would mean that a candidate fails the examination because of the large number of poor quality questions and not a lack of knowledge and/or understanding of the law. This danger far outweighs the problems that answering verification questions in the examination creates for an individual candidate.

One suggestion I have heard is that as the NPIA examination question bank must contain hundreds if not thousands of multiple-choice questions, the need to verify new questions is no longer required. Such an approach could be taken but it would mean that the examination question bank would remain static and be unable to ask questions on new law as it was introduced. In addition, as 'old' law fell away, so would the multiple-choice questions on that 'old' law in the examination and the NPIA question bank would offer less and less questions for examiners to choose from. Consequently, within a short period of time the examination would become stale and verification questions would have to be re-introduced.

A further point that a candidate may make is to ask why verification questions cannot be verified somewhere else and not within the Ospre Part I examination? This is because the only place that accurate feedback on multiple-choice questions can be obtained is in the actual examination. If NPIA were to offer candidates the facility of sitting through 50 verification questions, how many candidates' would take up the offer? Even if they did, there is no guarantee that the questions would be taken seriously and therefore the feedback could be called into doubt. Placing verification questions in the examination enables realistic and large-scale feedback to be obtained to ensure that the question can legitimately be introduced at a later stage having undergone a strict testing regime.

In conclusion, candidates would be far worse off if verification questions were not included as part of the examination process. They are an absolute necessity to ensure that questions are not only current but also fair (please see the 'Myth, Rumour and Barrack-room Advice' section for further information on verification questions).