Statistical consultancy · United Kingdom
Statistical consultancy for health behaviour-change programmes
Northwood Metrics is a statistical consultancy specialising in the evaluation of weight management, diabetes prevention and lifestyle behaviour-change programmes.
We design evaluations and surveys, model the data that services already produce, and review analyses carried out by others. Our clients are commercial and third-sector providers, commissioners and public bodies in the United Kingdom and the United States. Engagements run from a short review of an analysis already completed to a full evaluation carried through to peer-reviewed publication with the provider as co-authors.
What we do
Programme and service evaluation
What a programme achieves under real delivery conditions, from the records a service already holds.
Survey design and analysis
Sampling, weighting and non-response, and estimates that carry the design through to the standard errors.
Regression and multilevel modelling
The general workhorse: relationships between variables, estimated in a form that survives a sceptical read.
Panel and longitudinal analysis
Repeated measurements on the same people, and the change within them rather than the difference between them.
Survival and time-to-event analysis
Not whether something happened but when, with the people it has not happened to yet still in the analysis.
Study and experimental design
What to measure and how, agreed and dated before the data exists, because most later problems are created here.
Case studies
Dose–response in a group-delivered childhood programme
A provider running a group weight management programme for children across several hundred community sites asked whether attending more sessions related to a greater change in outcome, and from what level of attendance onwards. We fitted variance-component models to establish how much of the variation sat between groups rather than between children, then estimated the relationship with the between-group and within-group components separated. The shape was checked with a non-parametric fit before any linear form was assumed, and a specification allowing the relationship to change at a threshold was compared against one that did not.
Peer effects in a group-delivered service
The same provider had reason to think that the composition of a group mattered to the individual children in it. Everyone in a group shares a venue, a deliverer and a cohort, so ordinary regression cannot separate the influence of peers from the environment they have in common. We absorbed the shared component with group fixed effects and used an instrumental-variables strategy to identify the peer relationship, then carried the analysis through peer review with the provider as co-authors.
Whether outcome differences run through attendance
A provider delivering a diabetes prevention programme nationally knew that outcomes differed across age, deprivation and ethnicity, and needed to know whether those differences reflected a different response to the programme or a different amount of it received. We specified a mediation analysis with attendance as the candidate mediator, estimating total, direct and indirect paths with inference by bootstrap, and set out which parts of the pathway the design could speak to and which it could not.
Replicating a published relationship in a provider’s own data
An adult weight management provider wanted to know whether a relationship reported in the published literature held in their own records. We replicated the published selection procedure first, so that any difference could be attributed to the data rather than to the method, then re-estimated the same relationship on the full enrolled sample without conditioning on a variable measured after enrolment. Both specifications were reported side by side, with the difference between them traced to the sample each was estimated on.
Attendance, retention and the reported effect
Most reported programme effects describe the participants who completed. The people who never started, and those who left early, are absent from that figure by construction, and they do not leave at random.
Differences between groups in a reported outcome frequently reflect differences in how much of the programme each group received rather than differences in how the programme worked for them. The two have different operational answers, and neither is visible in a headline average, so the analysis separates them before either is reported.
Discuss a piece of work
The most useful first message says what the programme is, what data exists, and what date the work has to meet.