Chapter 2 The development of OASys User surveys Research has also examined the views of professionals regarding risk assessment instruments. The studies by Aubrey and Hough (1997), and Aye Maung and Hammond (2000 ) are particularly relevant. The former identified the needs of Probation officers when using Risk Assessment Scales, and the latter compared the usability and value of LSI-R and ACE in probation settings. Based on these two studies and feedback from the user group and those involved in the pilot study, we identified user requirements for the new assessment instrument as follows.
Face validity It is important that the assessment system is constructed so that the contributory items make sense to the assessor and the offender. It must be clear why each item is included. Only if the value of collecting information is readily recognised can we expect practitioners to conduct the assessments conscientiously. Clear definitions It is important that consistency exists between assessors. This can only be achieved by providing clear and unambiguous definitions of the items which comprise the instrument. Difficulties in this area were highlighted by the Aye-Maung and Hammond studies .
Of course no system can ever be completely prescriptive, there will always be specific instances where the individual offender does not exactly fit the framework. But it is important that clear concise and full definitions of each item are provided. OASys does this by providing prompts within the assessment itself, and a fuller account of the item in the manual. The manual also contains examples of where a judgement of 0, 1 or 2 is given, and guidance about what information should be taken into consideration.
Simple scoring system Aubrey and Hough’s study looked particularly at types of scoring systems and found that practitioners would rather have a simple scoring system, preferably one which required a straightforward binary scale of yes or no . Scales that stretched beyond five points were found to be difficult, as it was not possible to make judgements at such fine gradations. The OASys system uses a simple yes/no system where possible. In areas where practitioners need to make some judgement as to the degree of the problem, a simple 0, 1, 2 scale is used consistently. Totalling the scores involves weighting the various sections.
In the computerised version of OASys this will be done automatically, and in the paper-based version computation has been reduced to a minimum. OASys Manual V2 © Crown Copyright 2002 Page 9