Most universities arrive at programmatic assessment department by department. Medicine starts because a new programme director came from a school that ran it. Dentistry starts two years later, independently, because the ADEE framework pushed them there. Nursing has been doing something close to it for a decade under a different name and does not recognise the vocabulary. Each has built something that works locally.
Nobody made a bad decision at any point. The result is still a problem, and it becomes visible at exactly one moment: when the institution has to answer a single question about itself.
Three departments running programmatic assessment three ways is not local autonomy. It is three incompatible answers to one accreditation question, and the question is asked of the institution.
Where the divergence actually bites
Not, usually, in the day-to-day. Departments can run different instruments quite happily for years. It bites in five specific places.
Accreditation and institutional review
The AQU and ANECA seguiment and acreditació cycle asks the institution for evidence, and the six dimensions of the review do not vary by faculty. If medicine can produce competency coverage as a live matrix and dentistry produces a spreadsheet assembled in March from three sources, the institutional submission is only as strong as the weaker one. The strong department's work does not raise the average; the weak one lowers it.
The data protection position
Every separate assessment system is a separate processing activity: its own record in the register, its own risk profile, its own impact assessment where required, its own processor contract, its own retention schedule, and its own answer when a student exercises a subject access request. A DPO covering four faculties with four systems is maintaining four documentation sets for one category of processing, and is exposed at whichever one was set up with the least care.
This is the cost that is most consistently underestimated at the point where the decision to let each department choose is taken, because it lands on someone who was not in that meeting.
Student experience across a shared year
In institutions where students take shared modules or interprofessional placements, divergent systems mean a student can be assessed under two different models in the same term, with different portfolio expectations and different language for the same thing. Students notice this faster than faculty do, and it damages their confidence in the assessment regime generally.
Staff who work across programmes
Clinical supervisors frequently teach into more than one degree. Two logins, two vocabularies and two workflows is a reliable way to reduce the quality of what they record in both.
Institutional memory
A departmental system built around one enthusiastic academic is a succession risk. When they move, the institution often discovers that the configuration, the mappings and the thresholds existed largely as that person's working knowledge.
What "one model" should and should not mean
The failure mode in the other direction is real, and worse. An institution mandates a single configuration, dentistry is told to use instruments designed for medicine, clinical relevance drops, completion rates fall, and within two years the departments have rebuilt their own parallel spreadsheets alongside the official system. Now you have the original problem plus a procurement.
The distinction that holds up in practice is between what has to be shared and what has to be local.
Shared, institutionally:
- The data model. One definition of what an assessment record is, what a competency is, and how they relate.
- Identity and access. One SSO, one role model, one leavers process.
- The data protection position. One processor, one impact assessment, one retention policy, one answer to a subject access request.
- The audit and export format, because that is what the institutional submission is built from.
- Accessibility and language, which are institutional legal obligations rather than departmental preferences.
Local, to the department:
- The competency framework. Dentistry on the ADEE Graduating European Dentist, medicine on its own official set, nursing on its own; all mapped to official resultados de aprendizaje, none forced onto each other.
- The instruments in use, and which competencies each one addresses.
- Thresholds, minimum counts and what constitutes sufficient evidence, which are academic judgements belonging to the discipline.
- Committee composition and decision rules.
- Rollout pace. Departments do not have to start in the same term, and should not.
An institution that gets this split right can answer the accreditation question once, hold one data protection position, and still let a dental faculty run dentistry's assessment the way dentistry needs it run.
The Moodle question, answered before it is asked
The first question from institutional IT is always the same: which system is the system of record.
Our answer is fixed, and it is a constraint we impose on ourselves rather than a feature we might negotiate away. Final grading authority and transcript data residency remain with the institution's LMS — Moodle in most Spanish institutions, Blackboard in many UK ones. eLEA.to operates as an append-only competency tracking ledger alongside it, integrated over LTI 1.3 with SSO and roster sync, and it does not hold the official academic record.
We are aware that this is occasionally an unpopular answer in the room. We have had a medicine department ask whether eLEA.to could become the single source of truth across all their grading, precisely because their Moodle instance had accumulated years of inconsistently named subjects and they wanted out. It is a reasonable thing to want, and it is a different product decision with different regulatory consequences, so it belongs in a separate conversation rather than being conceded in a demo.
Honest costs of consolidating
Migration is the expensive part, and it is academic work. Mapping legacy competencies onto official learning outcomes under RD 822/2021 is not a data import. It requires someone with academic authority to decide the correspondences and sign them off, per degree. A platform can hold the mapping, version it and keep an audit trail of it. It cannot make the decisions, and a project plan that assumes otherwise will slip by a term.
The first department pays for the others. Whichever faculty goes first absorbs the configuration work, discovers the edge cases and makes the mistakes. This should be recognised institutionally rather than framed as a pilot they volunteered for.
Cultural change is slower than the software. Govaerts, Van der Vleuten and Schut describe organisational "deep structures" — communication patterns, competing values, taken-for-granted assumptions — that resist assessment reform independently of tooling. The Ottawa 2020 group reported that most implementation barriers related to the need for a paradigm shift in assessment culture rather than to systems. A platform is a necessary condition for institutional consistency and nowhere near a sufficient one.
You will find out things about your programmes. Consolidation makes comparison possible, and comparison between departments is politically live in a way that a departmental system never was. This is a real cost and it is worth naming in the business case rather than discovering in a faculty board.
The claim, narrowly
We are not claiming that one platform makes an institution's assessment coherent. Coherence is an academic governance achievement and it takes years. What a shared platform does is remove the reasons for incoherence that are purely technical: incompatible records, four data protection positions, an institutional submission limited by its weakest department, and a supervisor holding two logins for the same job.
That is a smaller claim than the sector usually makes. It is also the part we can actually deliver.
References
- Govaerts, M., Van der Vleuten, C. and Schut, S. (2022) "Implementation of programmatic assessment: challenges and lessons learned", Education Sciences 12(10), 717.
- Torre, D., Schuwirth, L., Van der Vleuten, C. et al. (2022) "An international study on the implementation of programmatic assessment", Medical Teacher 44(8), 928–937.
- Heeneman, S. et al. (2021), Medical Teacher 43(10), 1139–1148.
- Real Decreto 822/2021 and the AQU Catalunya Marc VSMA framework.