I guess there’s a timing issue I hadn’t thought of (I haven’t had the daily email round up of consultations yet today and I can’t honestly recall whether it arrives the same time every day or not) but a quick check of the gov.uk website shows only one consultation has been published today, or at least that was all there was when I started writing.
It’s a code of practice for Forensic Gait Analysis (FGA) published by the Forensic Science Regulator, an agency of the Home Office. I thought I’d just speed read it but actually it’s more interesting than I’d expected. I mean, “forensic gait analysis” is the pseudoscience of identifying people by the way they walk. As it says in the paragraph “limitations of service”:
- forensic gait analysis predominantly relies on observational analysis and comparison, rather than measurable objective techniques;
- features of gait have discriminatory potential but cannot currently be used to identify a person in isolation or from an open population;
which makes sense, because if you think about it you can identify your partner or your child or parent in a crowd without being able to see their face, just by the way they move. But if you notice someone with a distinctive red hat and then a few days later see someone in a similar hat, could you identify them as the same person from the way they walked? From examining video of the way they walked? I know I couldn’t – but then would I be able to if I trained as a Forensic Gait Analyst? And would you want to go to court reliant on my evidence if I did?
So I speed read the CoP and it looked reasonably sensible – I’d like to see someone with experience of the use of FGA in court read and comment on it, but no doubt legal networks are or will become aware of the consultation and be on it. The bit that worried me was:
17.1.4 Methods used by the Forensic Unit can be acquired by:-
i) developing a new method within the Forensic Unit;
OK, so you can develop an innovative way of undertaking FGA. But how do we know it works? Ah. You have to have it validated. And how do you have it validated?
17.2.1 When validating a method for use in forensic gait analysis the Forensic Unit shall:
i) determine the end-user’s requirements;
ii) determine the specification of the method;
iii) conduct a risk assessment of the method;
iv) review the end-user requirements and specification;
v) define the acceptance criteria for the method;
vi) produce a validation plan for the method;
vii) detail the outcomes of the validation exercise;
viii) independently assess the validation work to ensure the method complies with the acceptance criteria for the specification;
ix) produce a validation report;
x) produce a statement of validation completion; and
xi) devise a plan for implementation and monitoring of the method.
Is it just my nasty suspicious nature that reads this as “invent your own method, and then write a report that shows how well it works and how carefully you’re going to use it”? Although to be fair 17.2.3 goes on to say they expect most of the validation will be done by peer reviewed academic study.
When I skimmed to the end it finally clicked what I was looking at. This ought to be an internal manual: how to run this bit of your forensic department. It might be made public for FoI purposes but no-one except those affected would ever trouble to look at it. It needs to be a code of practice because we’ve outsourced this kind of work to profit-taking entities, people for whom the idea that they should have “facilities [that] cater for the safe storage of casefiles to maintain the integrity and identity of technical records” is not a given.
The kind of people, in fact, who need a forensic science regulator in the first place.
tiintax 