A practical guide to reading, writing and challenging psychological test results in expert evidence
Free, in English, 25 pages. For judges, prosecutors and lawyers who receive an expert opinion that rests on a psychological test, and for psychologists who write such opinions.
A psychological test in expert evidence usually arrives as one number: a score of 68, the 93rd percentile, “below the cut-off”. Behind the number sits a chain of decisions that a court rarely sees. The guide sets out that chain and puts ten questions that test it.
The ten questions
- Is the instrument valid in this language, for this purpose and this population?
- Which norms were used, how old are they and on whom were they collected?
- What exactly does the cut-off mean and where does it come from?
- What is the standard error of measurement, and is the score reported with a range?
- Has the person been tested before with this or a similar instrument?
- How were response style and possible malingering checked?
- Was the instrument legitimately used, and is the expert authorised to use it?
- What can the score not show?
- Where do facts end and legal conclusions begin?
- What must the file contain, and can the work be checked?
Each question comes with an explanation, what to ask, what a good answer looks like and the warning signs. The guide also has invented numerical examples, an invented opinion read through the ten questions, a one-page checklist and a glossary. It describes the measurement side of a test-based opinion, which is the same wherever the opinion is given. It does not describe any national law, is not legal advice and concerns no real case.
One-day course: test results in expert evidence
A practical course, built on the guide and on invented cases, for expert psychologists, lawyers, judges and prosecutors. A mixed group works best, because the same opinion is read by three professions and each asks different questions.
By the end of the day, participants can:
- judge whether an opinion names its instrument, version, language and norm, and whether the evidence cited supports the use made of the score;
- calculate a score interval from the standard deviation and reliability, and judge whether a score can be reliably separated from a cut-off;
- tell reliable change from measurement error and practice effects when scores from repeated testing are compared;
- appraise the checks of response style and effort, and see which conclusions about intent they cannot support;
- rewrite a conclusion so that it states measured facts and leaves legal qualification to the court;
- draw up a file checklist that lets a second qualified person repeat the scoring.
Programme (about six hours of teaching)
| Block | Topic |
|---|---|
| 1 | Instrument, language and norm: what an opinion must name |
| 2 | Measurement error and the cut-off |
| 3 | Repeated testing, response style and the limits of a score |
| 4 | Facts and legal conclusions |
| 5 | The file and how to check it: a workshop |
Each block has a short presentation, a worked example, an exercise and a check at the end. Up to 20 participants. Delivery format, dates and fees are agreed with the organiser; fees and dates on request.
Enquiries: info@centerforpsychology.me
Related service
If a court, a party or an expert needs an independent view of whether an assessment instrument is valid, or a response to another side’s expert report, see Expert witness. The guide and the course do not replace it.
About the author. Dr Milos Kankaras is a psychometrician with a PhD in social sciences (Tilburg University) and more than 25 years of research experience in developing and checking measurement instruments, including large international studies. He does not carry out clinical evaluations of individuals. He develops psychological instruments and provides measurement-quality services, including independent expert opinions on whether assessment instruments are valid, so readers should know of that interest.