The Real Examiner Sits in the Witness Box, Not the Lecture Theatre

Bottom line up front: A forensic science graduate's competence is not settled in the lecture theatre. It is tested in the witness box, where an opinion must survive cross-examination, and in the accredited laboratory, where a method must satisfy an international competence standard and, in England and Wales, a statutory Code of Practice. A generic end-of-module satisfaction score measures none of that. Worse, in a discipline with a well-documented mismatch between graduate numbers and forensic jobs, a high satisfaction mean can coexist with poor preparation for the only arenas that matter. Course evaluation for forensic science has to be built around the courtroom and the lab, not around the classroom.

A discipline governed by external standards, not opinion

Most subjects are quality-assured against academic benchmarks. Forensic science is quality-assured against those and against the machinery of the criminal justice system, and the two are not the same instrument.

Programmes are accredited by the Chartered Society of Forensic Sciences (CSFS), whose scheme is unusual and instructive: it accredits components, not whole degrees. Under its Component Standards model an institution must complete the core standard — Interpretation, Evaluation and Presentation of Evidence — plus at least one, preferably two, further standards drawn from Laboratory Analysis, Crime Scene Investigation, Forensic Anthropology, Forensic Archaeology or Digital Forensics. The standards are designed to align with the Quality Assurance Agency subject benchmark, and each accredited institution must nominate a University Quality Lead who is a professional member of the Society. The accreditation question is not "are students satisfied?" but "does this specific component equip a student for evidential work?"

Downstream, the laboratory is accredited to ISO/IEC 17025:2017 — "General requirements for the competence of testing and calibration laboratories" — with the United Kingdom Accreditation Service applying ISO/IEC 17020 to crime-scene examination and ISO/IEC 17025 to laboratory testing. The organising concept in both standards is competence, demonstrated and audited, not reported preference.

The statutory backstop: the Forensic Science Regulator's Code

Since the Forensic Science Regulator Act 2021, the Regulator has been on a statutory footing with power to issue a binding Code of Practice. The Code, dated March 2023, came into force at 00:01 on 2 October 2023 and applies to forensic work carried out on or after that date (Crown Prosecution Service guidance). It sets quality-standard requirements across a list of forensic science activities and obliges practitioners to declare compliance and, where they cannot, to address the competence of the practitioner, the validity of the method, its documentation, and the suitability of equipment and environment. Non-compliance does not automatically make evidence inadmissible — the court retains discretion — but it is now a statutory expectation that shapes what a competent graduate must be able to do. No other laboratory-science degree is taught in the shadow of a statutory Code of this kind.

Competence is proven under adversarial conditions

The forensic scientist's ultimate assessment is adversarial. In England and Wales, Part 19 of the Criminal Procedure Rules makes an expert's overriding duty to the court — a duty that "overrides any obligation to the person from whom the expert receives instructions or by whom the expert is paid" — and requires disclosure of anything that might undermine the opinion. The Law Commission's 2011 report Expert Evidence in Criminal Proceedings went further, criticising a "laissez-faire" judicial approach and recommending a statutory reliability test. Whether or not a graduate can write an intelligible, defensible report and hold an opinion together under cross-examination is the assessment the profession actually cares about. A five-point scale asking whether the seminars were engaging has nothing to say about it. This is the same "competence is validated after the degree, in a high-stakes external forum" pattern we have described for law programmes and the SQE and for sport and exercise science lab competency.

The graduate-outcomes problem the satisfaction mean hides

Forensic science carries an unusually sharp version of the employability question. As far back as 2005, the House of Commons Science and Technology Committee's Forensic Science on Trial noted that a UCAS search returned "a list of 401 degree courses at 57 universities," and heard that this expansion "does not reflect the limited employment prospects in forensic science." In 2019 the House of Lords Science and Technology Committee's Forensic science and the criminal justice system recommended that the Office for Students review forensic science courses, their quality and the job-prospects information given to applicants. When supply so far exceeds demand, a healthy satisfaction score is not reassurance — it may simply mean students enjoyed a course that did not prepare them for the narrow, exacting roles the field offers. Part of that gap is expectation: peer-reviewed work on the "CSI effect" documents how media portrayals create a disconnect between what students expect and what forensic practice demands (see, for example, Teaching To Avoid the "CSI Effect", Journal of Chemical Education, 2006). Evaluation should be surfacing that gap, not papering over it. We have made the parallel argument about satisfaction masking outcomes in employability and the skills gap.

But doesn't the student voice still matter?

The obvious objection: if competence is judged in the courtroom and the lab, why survey students at all? Two reasons. First, students are the only continuous witnesses to the learning environment — the quality of feedback, the realism of casework, the adequacy of lab access and supervision, the clarity of expectations — and those are precisely the conditions that produce (or fail to produce) court-ready graduates. Second, dismissing the student voice would replace one narrow instrument with another. The honest position is not "ignore students" but "stop asking students to certify their own competence, and start asking them the questions they are uniquely placed to answer." A first-year cannot tell you whether their evidential reasoning would survive cross-examination; they can tell you whether they ever practised it under realistic conditions.

What to measure instead

• Ask about court- and lab-readiness, not enjoyment: did students produce and defend expert-style reports, work to documented methods, and experience realistic adversarial challenge?
• Survey the practical environment: lab access, equipment, casework realism, supervision quality and the integrity culture around evidence handling.
• Read the free text for the expectation gap: the distance between why students enrolled ("CSI") and what the course revealed the work to be.
• Connect feedback to outcomes honestly: pair evaluation with candid destination and job-prospects information rather than letting a satisfaction mean stand in for either.

Where Koji fits

The static, average-a-Likert-score survey is exactly the wrong instrument for a discipline judged under adversarial and accredited conditions. Koji for Education uses AI-moderated conversational interviews that can probe whether students actually practised evidential reasoning, worked to documented methods, and were tested under realistic challenge — going beyond a number to the reasoning behind it. Its automatic thematic analysis surfaces the expectation gap and the environment concerns buried in open text at scale, and its standardized, bias-aware moderation keeps questioning consistent across cohorts where a human interviewer would not. Programme-level reporting lets a course team connect what students say to what the CSFS component standards and the labour market require. Koji mitigates and surfaces the signals a satisfaction mean hides; it does not claim to replace the professional and judicial judgement that ultimately certifies a forensic scientist. The same AI interview engine underpins general user and customer research on the main koji.so platform.

The forensic science graduate is examined, in the end, by a barrister and by an accreditation auditor. An evaluation instrument that ignores both is measuring the wrong room.

Frequently asked questions

Why is a standard satisfaction survey a poor fit for forensic science? Because the discipline's real assessments happen in the courtroom, where an opinion must survive cross-examination, and in the ISO 17025 accredited laboratory, under a statutory Code of Practice — none of which a Likert satisfaction score measures. A high mean can coexist with poor preparation for those arenas.

What is distinctive about CSFS accreditation? The Chartered Society of Forensic Sciences accredits components, not whole degrees. Institutions must satisfy the core Interpretation, Evaluation and Presentation of Evidence standard plus further standards such as Laboratory Analysis or Crime Scene Investigation, each aligned to the QAA subject benchmark.

What is the Forensic Science Regulator's Code of Practice? A statutory Code, empowered by the Forensic Science Regulator Act 2021 and in force since 2 October 2023, setting quality-standard requirements for forensic science activities in the criminal justice system. It obliges practitioners to declare compliance on method validity, documentation, competence, equipment and environment.

If competence is judged in court, why survey students at all? Because students are the only continuous witnesses to the learning environment — feedback quality, casework realism, lab access, supervision and expectations — which are the conditions that produce court-ready graduates. The fix is to stop asking students to certify their own competence and start asking what they are uniquely placed to observe.

What is the "CSI effect" and why does it matter for evaluation? It is the documented gap between media-driven student expectations of forensic work and the reality of the discipline. Evaluation should surface that gap — through open-text analysis of why students enrolled versus what they found — rather than let a satisfaction mean obscure it.

How does Koji help with a discipline this specialised? By replacing a static survey with AI-moderated interviews that probe court- and lab-readiness and the practical environment, reading open-text signal at scale through thematic analysis, and reporting at programme level so a course team can connect student experience to accreditation and labour-market realities — while being explicit that it mitigates and surfaces rather than eliminates.

---

Evaluating a forensic science programme and want feedback aimed at court- and lab-readiness, not just satisfaction? See how Koji for Education works.