'Psy’ expert evidence in the family courts: The potential for corpus-assisted analysis

This article introduces corpus-assisted linguistic methods as an exploratory means of analysing expert psychologists’ reports used in public family law (child protection) cases. Analysis of this dataset is a new application for corpus linguistics (CL) and the primary purpose of this article is to explore viability and potential for its future research using CL as a core method. For this study we have created and analysed a 25 single-text-type specialised written corpus consisting of 25 expert psychologists’ reports (the Psychology Report Corpus “PRC-25”). The reports are a random sample selected from a population of all psychologists’ reports held in Cafcass files over a 10-year period, representing the first corpus of its kind in a currently under-researched area. Our study uses both an inductive (data-driven) approach to identify significant themes and topics in the reports, and a deductive (legal-intuitive) approach to explore psychologists’ use of legally significant terms, especially risk of and significant harm. We also explore the possibility for using this new methodological protocol to triangulate analysis of a larger and representative corpus of expert psychologists’ reports, and the possibilities for corpus-driven analysis of the genre of written expert evidence text types more generally.