Standard or individualized measures

There are many standard measures already in existence, especially ratings and questionnaires, covering most areas of clinical interest. However, the fact that a scale is well used does not necessarily mean that it is psychometrically sound, (318) or that the content covered is appropriate for a given clinical or research purpose. Rating scales may be widely used—such as the long-established NOSIE scale published in 1965—even when they are no longer the best scales technically, but because the volume of published research using them permits more comparisons to be made with other studies. Regier et a/.(13 point out the need to standardize measures for both clinical and epidemiological work, given the 'drift' or 'mutations' that can occur with even the most carefully designed measures, with major consequences for public health and policy if prevalence estimates cannot be made reliably.

One sound solution to this dilemma is to create a totally new measure, making sure it is better than its predecessor, or systematically to improve the properties of an existing measure further. Parker et a/.(29 describe a modification of the well-established Parental Bonding Instrument to include abusive parenting, which was omitted in the original version. Their article demonstrates both how to modify an original measure to improve item wording, and at the same time incorporate additional material to increase the value of the measure. An unsound solution is to change the measure in an ad hoc way—scale vandalization!—with no awareness of the principles of sound scale and item construction, this only results in an instrument that will then be of poor or unknown reliability or validity.

Direct observation methods can be used in a standardized manner for a group of patients. But they do lend themselves to flexible modification, so that the frequency of observation, and the duration of observation periods, can be chosen to suit the requirements of the task. The most common problems encountered in designing or modifying behavioural measures are as follows: choosing coding categories which cover all the most likely events and are ranked appropriately, using clear and unambiguous language, and using exact frequencies (such as 'twice a day' or 'at least every hour') rather than vague terms such as 'often' or 'frequently'.

Mirror Madness

Mirror Madness

Get All The Support And Guidance You Need To Permanently STOP Having A Bad Body Image. This Book Is One Of The Most Valuable Resources In The World When It Comes To A Guide To Better Body Image.

Get My Free Ebook


Post a comment