Research and analysis services
Designed, recorded, and reported so that findings can be relied upon by the people who commissioned them.
Research Programme Design
Study design, sampling strategy, and the written analysis plan, agreed and dated before any data is collected or examined.
Survey and Field Research
Instrument design, cognitive testing, and fieldwork across English and Spanish-speaking populations, with both language versions developed in parallel.
Programme Evaluation
Assessment of whether an intervention achieved what it set out to, including where the finding is that it did not.
Statistical Analysis
Analysis executed against the recorded plan, with any departure documented, explained, and reported alongside the result.
Data Quality Assessment
An honest account of whether the records you hold can answer the question you are asking of them, before anyone builds on them.
Reporting and Briefing
Written findings and in-person briefing for boards, funders, and programme staff, in language they can act on without a statistician present.
What gets recorded before we start
The questions
Stated specifically enough to be answerable, and few enough to be answered. A plan listing twenty questions is a plan to go fishing with paperwork attached.
The measures
Exactly how each concept will be operationalised, including which items compose a scale and how missing responses will be treated.
Comparison groups
Who is being compared to whom, defined before anyone knows how the groups will perform.
Sample and power
How many responses are required for the planned comparison to mean anything, and what will happen if fieldwork falls short.
Interpretation rules
What result would count as supporting the hypothesis and what would count against it, agreed while both outcomes are still live possibilities.
Handling of subgroups
Which subgroup analyses are planned, and an explicit acknowledgement that any others will be reported as exploratory.
From question to briefing
Framing. We start with what decision the findings will inform and who will be reading them. A study designed for a board briefing and one designed for a funder report are not the same study, and establishing that first prevents an expensive redesign later.
Feasibility. Before design work begins we establish whether the question can be answered at all with the records available, the population accessible, and the timescale you are working to. Where it cannot, we say so and that is the end of the engagement.
Design and plan. Sampling strategy, instruments, and the written analysis plan, developed with you and dated. This is the phase most programmes underinvest in and the one that determines whether the findings will hold up.
Fieldwork. Instruments are cognitively tested in both languages before deployment, interviewers are trained and observed, and response patterns are monitored while collection is live so that problems are caught during fieldwork rather than after it.
Analysis and reporting. Executed against the recorded plan, with confirmatory and exploratory findings separated, departures documented, and results briefed in person to the people who will act on them.
Handover. Instruments, cleaned datasets, analysis code, and the dated plan transfer to you at the close of the programme. A different team should be able to reproduce what we did without asking us anything, and that is the standard we build the documentation to.