Forensic Psychology For Dummies (90 page)
Read Forensic Psychology For Dummies Online
Authors: David Canter
Different questionnaires have been developed to help in the assignment of people to the different diagnostic categories (including the MCMI and PAI ones listed in the earlier Table 9-1).
All the authorities who produced these classification schemes emphasise that the DSM and ICD systems are guidelines that can be used only by clinically trained individuals. They aren’t ‘cookbooks’ to be followed without carefully guided experience. This salutary warning indicates that diagnosis of these mental disorders is more of a craft than an objective scientific procedure.
The five DSM axes of mental disorders
DSM identifies what it calls five ‘axes’ of mental disorders, each containing descriptions or definitions of particular mental problems. Forensic psychologists most frequently draw on axes I and II:
Axis I:
These are the disorders that bring people into a mental health clinic, such as major mental problems like schizophrenia, drug addiction and other forms of substance abuse disorders. They also include severe depression, anxiety disorders, bipolar disorder (which used to be called ‘manic-depression’). The conditions typically identified in children such as attention deficit hyperactivity disorder (ADHD) and autism spectrum disorders are axis I as well. Eating disorders, notably anorexia nervosa and bulimia nervosa also belong in this axis.
These are the aspects of personality and intellectual disabilities. This includes being paranoid, schizoid and the other personality disorders that I consider in Chapter 10.
Acute medical conditions and physical disorders.
Psychosocial and environmental factors that contribute to the disorder.
Global assessment of functioning, which includes how the person copes with the challenges of daily life.
Testing the Tests
Obviously, experts need to know how effective the different psychological tests are, because they vary enormously. To this end, a number of characteristics of tests have been identified that give an indication of their qualities. Understanding a test’s good and bad points is essential because it helps you to evaluate how valuable a test is likely to be and what weight you can put on its results.
Imagine a ruler that’s made out of very flexible elastic for measuring length, which gives you a different result every time you use it for measuring the same piece of metal – that’s not a measuring tool you’re going to trust again.
Aiming for test reliability
The most basic quality any assessment instrument must have is a high degree of reliability.
is the likelihood that carrying out the same test under very similar conditions, on more than one occasion, gives the same results.
The establishment of reliability is more difficult for psychological measurements than for measures of physical objects, but broadly the same process is used – the test is given under similar conditions to the same people on different occasions to see how close the measurements are to each other. Of course people change more than a lump of wood does and they may even learn something from carrying out the test the first time; and so various ways around this have been devised, such as having two very similar tests administered at the same time.
In general, perfect reliability is never expected. A measure that varies between 0.0 and 1.0 is used to assess reliability. Anything above 0.9 is regarded as excellent, but tests that achieve around 0.8 are in common use, even reliabilities as low as 0.6 aren’t unusual.
Evaluating a test’s validity
A test can be very reliable and produce consistent results (see the preceding section) and yet still not really measure what it claims to measure. For example, a thermometer gives you a reliable measure of temperature, but isn’t very accurate if you used it to measure altitude! The problem with testing psychological characteristics is that (unlike physical objects) determining what you’re actually measuring isn’t easy. A measure that claims to be assessing how authoritarian people are, for example, may just be measuring their conventionality.
validity
.