Selected programmes
Described without identification of clients, in accordance with our engagement terms. Two produced null results.
An evaluation that found no effect
A community health intervention assessed against pre-recorded questions. The programme produced no measurable change, and we reported that rather than searching for a subgroup in which it appeared to work.
A plan signed before fieldwork
A foundation study where the analysis plan was agreed and dated before a single response was collected, which removed any later argument that the questions had been chosen to suit the answers.
An exploratory finding, labelled
A striking pattern noticed during analysis that had not been planned for. Reported as a hypothesis for future testing rather than as a conclusion, and subsequently tested properly in a second study.
A departure from the plan, documented
Fieldwork that fell short of the required sample. The change in approach and its consequences for interpretation were recorded in the report rather than quietly absorbed.
A question the data could not answer
Administrative records lacking the granularity the question required. We reported that at the assessment stage and proposed what collecting it would involve.
A bilingual instrument redesigned
A survey whose Spanish version was measuring something different to its English counterpart. Both were rebuilt in parallel and cognitively tested before fieldwork.
Why they are worth commissioning
A finding that an intervention did not work is frequently treated as a failed evaluation. It is not. It is the evaluation working, and it is often the most valuable result a programme can produce.
Continuing to fund something ineffective is the expensive outcome. An intervention that produces no measurable change will keep consuming budget indefinitely unless somebody measures it properly and reports what they find.
The pressure not to report it is real. Funders have committed money, programme staff have committed years, and everyone involved would prefer a positive finding. That pressure is exactly why the analysis plan is recorded in advance.
A null result is not the same as no information. It establishes an upper bound on the effect, identifies where the theory of change broke down, and frequently points at a more promising version of the intervention.
We help clients present null findings to funders. Delivered with the design quality that produced them, a null result reads as rigour rather than failure, and in our experience funders respond to it accordingly.
What clients ask before commissioning
What is in the analysis plan?
Questions, measures, comparison groups, sample requirements, interpretation rules, and which subgroup analyses are planned. Agreed with you and dated.
What if the result is null?
It is reported as the finding, with the same prominence as a positive result would receive.
Can you still explore the data?
Yes, and we do. Those findings are reported separately and labelled as exploratory hypotheses rather than conclusions.
What if you need to change the plan?
The departure and its reason are documented in the report, alongside what it means for how the result should be read.
Do you work in Spanish?
Yes, with instruments designed and tested in both languages in parallel rather than translated afterwards.
What happens to personal data?
Collection is minimised, retention is agreed in writing, and deletion is confirmed to you at the close of the programme.
Who owns the data and the findings?
You do. Instruments, cleaned datasets, code, and the analysis plan transfer to you at the close of the programme, so a different team could reproduce the work.
How long does a programme take?
Design and feasibility run four to six weeks. Fieldwork depends entirely on the population and the method. Analysis and reporting are typically three to four weeks after collection closes.