Psychometrics in everyday psychological practice

Bill Warren, Deputy President - Psychology Council

Questionnaires, tests, inventories, instruments, and so forth – let’s call them “psychometric instruments”, or simply “instruments” for present purposes,  can be useful when working with clients. Their use might be seen as something that differentiates us from our psychiatrist colleagues who, by and large, do not have specific training in their use nor the background knowledge of the research methodologies that generated them. The latter appear in research reports with titles like “The Construction and Validation of an Instrument to Measure …” – and there are many such. Some of these questionnaires appear only in the literature and are free; others are marketed commercially. Random examples of the former are the Locus of Control Scale and the Vulnerable Attachment Style Questionnaire, and of the second: the DASS and the Beck Scales (for depression and anxiety), which are widely used following purchase.

A traditional differentiation in relation to the use of these instruments is between:

  1. a treating professional who responds to questions by a relevant third party (e.g. a referring GP) in relation to their work with a client they have been seeing, or have seen in the past, for therapy under one or other of the different modes of referral: for example, under a Mental Health Care Plan, a Victims Counselling Approval, workplace or motor vehicle injury matter, DVA or VVCS, where pre- and post- measures are useful;  and
  2. an objective assessment and reporting for legal and related purposes by a practitioner who sees the client for the first time and usually only once, for the specific purpose of preparing a report. The context of the latter is of the practitioner being contracted, and hopefully accepted in the relevant context, as an expert witness. This status has to be earned and will be open to challenge in relevant contexts, for example under cross-examination in court.

There are two primary types of psychometric instruments available to a practitioner and these, more or less, ‘fit’ into the latter context. One is a range of what is perhaps best called “screening instruments” the practitioner might use to provide scaling of the problem(s) the client brings to him or her, indications of improvement as treatment proceeds, or simple ways of developing the client’s understanding of their problem(s), dilemmas, ‘blind spots’ and the like.

The second type is the well researched and validated substantive instruments with developed norms and built-in validity indices and/or sophisticated structural analyses, including  such matters as ‘faking good’ or ‘faking bad’. The first type may have sound methodology in their construction and validation but seldom have, for example, the same level of sample size or range as do the second. Examples of the second type ‘omnibus’ instruments are the MMPI, the PAI and the Millon Clinical Multiaxial Inventory when psychopathology is in question – and even the Rorschach is still mentioned in at least one psychology assessment handbook! Where psychopathology is not the primary focus then examples include the NEO Personality Inventory or more focused instruments like the MEIA (Multidimensional Emotional Intelligence Assessment) and the MEIA-W.

The reason for the present article is that the Council has had occasion to read both client files and formal reports where screening tests have been used and discussed as if they were more substantial psychometric instruments. Formal diagnoses have also been made on the basis of the client’s self-report on a screening instrument. The message here is simple: the utility of the information harvested by your use of psychometric instruments is crucially related to the empirical robustness of the instrument itself. It is cautioned that, when you consider using a screening device, you familiarise yourself with the conceptualisation of the concept or construct allegedly being measured by that particular “screener”. For example, the characterizations of Depression, Anxiety and Stress on the DASS are not the same as those used by other scales or by other theoretical perspectives.