What went well
The clearest achievement this week was closing out Objective 1: we finalised our research question — 'What mechanisms do parents/carers and students identify as building or eroding relational trust between disadvantaged families and their school within a Multi-Academy Trust?' Getting from an open, brief to a single testable question felt like real progress, and I think the organic/contractual/relational distinction WE'D been pushing for did end up doing some of the work of disciplining the scope — it gave us a way to say what the question is not asking (it's not asking about MAT governance legitimacy, it's asking about lived relational experience) as much as what it is. Alongside that, we also did substantial methodological groundwork: a session on mixed-methods research (worldviews, ontology/epistemology, qual vs quant, integration vs multi-method) gave us a shared vocabulary for describing what we're doing, and pushed us to be explicit that we're not just running qualitative and quantitative strands in parallel — we need actual integration (e.g. using attendance/pupil premium data to identify cases for follow-up interview, or using interview material to explain patterns in survey data), or we're just doing multi-method work and dressing it up. We also produced a proposal document with a full six-method approach (semi-structured and written surveys, existing trust/school documentation, numerical-scale surveys, attendance and demographic data) and a visual flowchart mapping the article structure end to end — from definitions through case study through literature review to recommendations — which gives the group a shared reference point going into the school-facing phase.
What could have gone differently
The research question is strong, but I don't think we've fully resolved the tension I flagged last week between what's theoretically central (relational mechanisms, told through families' own accounts) and what's practically accessible (documentary and quantitative data, given the access constraints). The six methods listed in the proposal include two qualitative interview/survey routes with students and parents directly, but those are still contingent on access we haven't confirmed — so on paper the design looks more qual-heavy and relational than what we may actually be able to deliver. I'd have liked to pressure-test the methods list against the access reality from the Cheney meeting before finalising it, rather than writing an aspirational toolkit and hoping the access conversation catches up.
Reflection and sense-making
The mixed-methods session was useful precisely because it gave me language for something that had been bothering me without my being able to name it: the difference between a multi-method design and a mixed-methods one. Looking back at our proposal, most of what we've specified so far is a list of methods sitting side by side, not yet integrated — we haven't said how the quantitative attendance/demographic data will inform or be explained by the qualitative interviews, just that we'll have both. That's not a flaw exactly, this is normal at proposal stage, but it clarified for me that the real intellectual work of the project is still ahead of us: deciding whether we're explanatory sequential (quant first, then qual to explain surprising patterns) or exploratory sequential (qual first, to build something we then test), because that choice will determine field order and probably which access battles are worth fighting first. Reflecting on the flowchart exercise too — mapping "Establish definitions → Institutional vs Relational Trust → Case Study → Lit Review → How does RLT influence schools/families → Suggestions or Looking forward → Limitations" — I noticed the structure is still quite linear/additive rather than integrative, which is the same pattern from the methods list.
Adjustment and development for next week
Behavioural adjustment: when we're specifying methods as a group, I'll push us to say explicitly which strand feeds into which (dominant/sequential vs equal-status/concurrent) rather than listing methods as a menu — this is the integration test from this week's session, and it's cheap to apply now and expensive to retrofit later.
Action for next week: I'll draft a one-page mapping of our six proposed methods onto an explicit mixed-methods design (most likely explanatory or exploratory sequential, given the access constraints mean quant/documentary data is more secure than qual access right now), so the group can decide with Objective 1 closed whether the design should be quant-led or qual-led — and I'll bring this alongside confirmation from the school on what qualitative access is actually realistic, so the design isn't built on an assumption we haven't tested.