Skip to main content

Diaposi officially launches 2 September 2026.

Diaposi logoBook a demo

Research & Development · Open to partners

Every lecture is
an unrun experiment.

Diaposi is a measurement instrument before it is a product. This page is addressed to the people who might want to point it at a real question.

Written by Alae Belaich, Founder & Developer · Last updated

Learning analyticsCognitive engagementNo biometricsGDPR-native

Contents

What is on
this page.

Fifteen sections, roughly twenty minutes end to end. Jump to whichever one you came for.

The question

Quantified learning and interaction.

A lecture hall generates an enormous amount of information every minute and throws essentially all of it away. Whether a concept landed, who fell behind and on which slide, which question half the room was privately holding. All of it evaporates the moment the session ends. Teaching remains one of the few skilled professions practised almost entirely without instrumentation.

Diaposi exists to capture that information. The research question is whether what it captures is actually worth anything.

Stated formally

Can volitional, slide-level interaction signals, captured without biometrics and without student accounts, serve as a valid real-time proxy for cognitive engagement? And does exposing those signals to the teacher mid-lecture change what students actually learn?

Two halves, deliberately. The first is a measurement question and it is answerable with correlational work. The second is a causal question about intervention, and it needs a trial. We are looking for partners on both.

Before anything else

A layer, not a replacement.

Research pages tend to make teachers nervous, and with good reason. Most sentences written about measuring classrooms are really sentences about measuring the people in them. So before the tables and the study designs, the part that matters most to the two groups who actually have to live with this.

Diaposi does not ask anyone to teach differently. You upload the deck you already made: the same PDF, the same order, the same board work, the same voice, the same pace. It adds a channel back from the room. That is the entire intervention. Nobody is required to gamify anything, adopt a new pedagogy, or restructure a course that has worked for twenty years.

If you have taught the same way for twenty years

  • Your existing slides work as they are. Nothing is rebuilt inside a tool.
  • You are never scored, ranked, or compared to a colleague. No number about you leaves your session.
  • You can ignore every signal on screen and the lecture runs exactly as it always did.
  • It is not attendance software. It cannot tell you who is in the room, because it does not know who anyone is.
  • Thirty seconds from upload to a live session. If it costs you a lesson plan, it has already failed.

If you would rather not be watched

  • No account, no install, no name. You join with a code and you are a participant, not a profile.
  • Your notes and your annotations stay in your own browser. There is no server that receives them.
  • Saying nothing is a valid answer. Below a quorum the system reports counts and passes no verdict at all.
  • Nothing here is graded. There is no participation score feeding into anything.
  • The screen shows the room, never a person. A question is a question, not a name attached to one.

A teacher who ignores every number on the screen is still using Diaposi correctly.

For a research partner this is not a caveat we tolerate. It is a design constraint we chose, and it has methodological consequences. An instrument that teachers resent is an instrument that gets switched off in week three, and a study with no week four is not a study. Consent, low friction and the right to stay silent are what make a multi-site, multi-semester dataset possible at all.

The instrument

What a Diaposi session already records.

None of this is a roadmap. Every signal below is captured today, per slide and per student, in sessions running in production, though two of them, deliberately, are captured only on the student's own device and never reach us at all. It is the part of the proposition that does not depend on anyone believing us.

SignalRecorded asWhat it means for research
Dwell time per slidesync / async / view / viewersTime-on-task, and the divergence between a student's pace and the teacher's.
Teacher time per slideprofSlideTimesThe intended pacing baseline every student signal is measured against.
Follow ratioavgAttention ÷ profTimeAttention alignment, measured against the teacher's own pace rather than an absolute standard.
Comprehension votelikes / dislikes per slideSelf-reported understanding, timestamped at the moment of instruction.
Signal strengthreviewShare, voterShareHow strong the verdict is, and how much of the room actually weighed in.
Questionscomments per slide, with peer upvotesConstructive and interactive engagement, anchored to a specific slide.
Hands raisedtotalHandsRaisedHelp-seeking behaviour, stripped of its usual social cost.
Pollsvotes, correctValue, textAnswersIn-the-moment formative assessment, including free text.
Expectations phaseexpectation → met / partial / not yetA stated learning goal before the session, self-rated after it. One pre/post pair per student.
Session modalityin_class / online / hybridA comparative condition recorded on every session. A natural experiment, already running.
Slide annotationOn devicepen / highlight / eraser, per slideWhere the material got marked up: by the teacher on the shared board, and by each student on their own copy. Density and location, never content.
Private notesOn deviceper-slide note list, timestampedWhat a student chose to write down, and at which moment of the lecture. A count could one day be contributed with consent. The text never.

On devicemarks the two signals that never reach a server. Notes and annotations live in the student's own browser storage, keyed to the session and the slide. There is no endpoint that receives them. We could not read them if we wanted to, and any future research use would be a consented, counted contribution rather than a collection.

Design decision 01

The engagement score is decomposed, not delivered.

Participation out of 40, understanding out of 35, interaction intensity out of 25, each shown with the counts it was computed from. A teacher can reconstruct the number by hand. There is no model in the middle, and nothing to audit that isn't already on screen.

Design decision 02

Below a quorum, the system refuses to judge.

A slide verdict requires at least three votes and at least a quarter of the students who reached that slide. Under that threshold Diaposi reports the raw counts and passes no verdict at all. Two dislikes out of three in a room of twenty is not “reteach this”. It is three people, and the other seventeen said nothing.

Method

What should be measured, and how.

Recording something is easy. Deciding what deserves to be recorded, at what grain, and what may never be concluded from it: that is the part with consequences. These six rules are already implemented in the product, and we would defend every one of them in review.

01

Measure the act, not the person.

Every row in the table above is an event a student chose to produce. None of it is an attribute inferred about who they are. The difference sounds philosophical and is entirely practical: acts can be counted, aggregated and forgotten. Inferred attributes follow people around.

02

Measure at the grain of instruction.

The unit is the slide, because the slide is the unit a teacher can actually act on. A session-level engagement score tells you a lecture went badly. A slide-level one tells you which four minutes to rebuild before next year.

03

Measure with a quorum, or say nothing.

At least three votes and at least a quarter of the students who reached the slide. Under that, Diaposi prints the counts and refuses to draw a conclusion. Most dashboards in this field will happily render a verdict from n = 2.

04

Measure what they did and what they said.

Dwell traces are temporally proximal and hard to fake, but semantically thin: a slide left open is not a slide being read. Taps are meaningful but self-reported. Neither is sufficient. Holding both on one timeline, in one session, is the methodological point of the whole instrument.

05

Never measure the body.

No face, no voice, no gaze, no heart rate. Not because it is technically hard, since it is not, but because it changes what a classroom is, and because European law has now agreed. The ceiling on accuracy this imposes is a price we think is obviously worth paying.

06

Measure the after, not only the during.

The live number is a trigger, not a result. What a teacher changes in the next delivery, and whether the next cohort does better, is the only outcome that settles anything. Any study built only on in-session metrics is measuring the thermometer.

Theory

The ladder these signals climb.

Chi and Wylie's ICAP framework ranks overt learning behaviours into four modes and predicts that learning increases as students move up them, by around eight to ten percent per rung in their original study. It is one of the most cited engagement frameworks in the field, and it is almost never instrumented directly, because measuring which rung a student is on usually requires an observer in the room with a coding sheet.

Mode 1

Passive

The learner only receives information.

Instrumented by

Dwell time and follow ratio. The slide was on screen, and we know for how long relative to the teacher's own pace.

Mode 2

Active

The learner manipulates the material.

Instrumented by

A deliberate tap on “Got it” or “Review”, which is a claim about their own understanding of this slide.

Mode 3

Constructive

The learner generates output beyond what was given.

Instrumented by

Writing a question, answering an open poll, stating a learning expectation before the session starts.

Mode 4

Interactive

Knowledge is built between people.

Instrumented by

Upvoting a peer's question, and the teacher re-teaching a slide because the room said so.

Diaposi instruments all four ICAP modes, live, at slide resolution, in an ordinary lecture hall with no observer and no setup.

The regulatory position

The state of the art just became illegal to deploy.

Multimodal learning analytics is where this field has been heading: cameras, eye trackers, microphones, wearables, heart-rate variability, fused into engagement classifiers reaching roughly 82% accuracy. It is genuinely impressive work. In a European classroom, most of it can no longer be used.

The EU AI Act classifies learning analytics in education as high-risk, and separately prohibits emotion-inference systems in educational settings outright: any attempt to read a student's emotional state from biometric data. That is not a compliance detail to work around. It closes the road.

Sensor-based engagement detection

Inferred from the body.

  • Facial video, gaze tracking, voice, heart-rate variability
  • Emotion inference, prohibited in education under the AI Act
  • Biometric data, so consent machinery for every participant
  • An instrumented room: thirty students, one site, one language
  • Runs after the fact, on a dashboard nobody sees during the lecture

The Diaposi approach

Declared by the student.

  • Every signal is volitional: a student chooses to send it
  • No face, no voice, no pulse, no emotion inference anywhere
  • No student accounts, anonymous by default, GDPR data minimisation
  • Any room with a browser: thousands of students, ten countries, one semester
  • Closed loop: the teacher acts on the signal during the lecture that produced it

We are not claiming the richest instrument. We are claiming the best one that is legal to run at scale in a European classroom, and scale is what this field's evidence base is actually missing.

The gap

The problem with the evidence is methodological, not conceptual.

Audience response systems have been studied for two decades. The meta-analyses report a small-to-moderate effect on cognitive outcomes (Hedges' g ≈ 0.47) and a moderate effect on non-cognitive ones (≈ 0.66). Encouraging numbers, except that the systematic reviews built on top of them are unusually blunt about their own foundations: weak variable control, few validated instruments, non-randomised designs that inflate effects, and no consensus on how these tools contribute to performance at all.

The reviews of multimodal learning analytics say something adjacent: the literature is largely offline dashboards and proofs of concept, with very little work on closed-loop systems running in real university environments.

Read together, the field is asking for exactly one thing: better designs running at greater scale in genuine classrooms. That is a tooling problem before it is a research problem, and it happens to be the problem Diaposi was already built to solve.

Open programme

Five studies we would like someone to run.

These are proposals, not results. Each one is scoped so that a single motivated research group could execute it inside one academic year, and each names honestly what we can contribute and what we cannot.

S1

Instrument validation

Do slide-level signals predict end-of-unit assessment beyond what attendance alone explains?

You bring

One faculty, one semester, access to assessment outcomes.

We bring

The full signal stream, per slide and per student, plus raw export.

S2

The closed-loop trial

Randomise sections: the teacher sees live slide status, or does not. Does the feedback loop itself move outcomes?

You bring

Randomisation design, ethics approval, assessment instrument.

We bring

Feature-flagged builds, reteach latency, per-slide intervention traces.

S3

The validity triangle

How far apart are dwell traces, self-reported comprehension, and graded outcomes, and when do they diverge?

You bring

Psychometric expertise and a validated comparison instrument.

We bring

Trace and self-report captured natively, on the same timeline, same session.

S4

Modality comparison

Do in-person, online and hybrid rooms produce structurally different engagement signatures?

You bring

Courses running in more than one modality.

We bring

Modality is already a field on every session. No new instrumentation needed.

S5

Voice and equity

Does anonymity change who participates, not just how much? Compare question authorship against hand-raising baselines.

You bring

Cohort demographics under an approved protocol.

We bring

Authorship distribution across anonymous questions and upvotes.

The after

Most of the value arrives once everyone has left the room.

The live pulse gets the attention, but it is only the trigger. What a session actually produces is a record: of what landed, what did not, what was asked and by how many people at once. That record is the part that changes the next lecture, and the next cohort's experience of it.

It is also where the two people this is really for finally get something back.

For the teacher who feels they are losing the room

A feeling becomes three slide numbers.

Teachers who sense their engagement slipping almost always already know. What they lack is where. “The second half went badly” is not actionable; “slides 14, 15 and 22 were flagged by a third of the room, and attention collapsed to 31% of your pace on slide 15” is a task for Sunday evening.

It is also, quietly, a defence. A teacher whose course gets questioned has evidence of what they adjusted and why, something the profession has historically had to argue without.

For the student who never speaks

Being heard without being looked at.

An anonymous question that collects twelve upvotes is a student discovering that the thing they were too unsure to ask was the thing half the room needed. That is a different experience of a lecture hall than the one they came in with, and it costs them none of the social risk that kept their hand down.

Afterwards their own recap holds what they contributed, what they flagged as unclear, and the notes they took, in one place, on their device, for them. Not a score. A record of having actually been there.

For research, this is where the real dependent variable lives. The interesting question was never “did the engagement number go up”. It is whether the teacher changed the next delivery, and whether the cohort that got the changed version did better.

Which is also why the session report exports as rows rather than a picture. The after is only useful if someone can compare it to the next one.

Limitations

What we don't know yet.

This section exists because the rest of the page is worthless without it. Every claim above is about what the instrument records. None of it is a claim about learning outcomes, and we are not going to borrow someone else's effect size and print it next to our logo.

A tap is a claim, not a measurement.

When a student marks a slide “Got it”, that is a statement about their confidence. Deslauriers and colleagues showed in PNAS that actual learning and the feeling of learning can be strongly anticorrelated: students in active classrooms learn more and believe they learned less. Our comprehension signal is exposed to exactly that failure mode. Quantifying the gap is study S3, not a footnote.

We have no longitudinal attainment data.

Nothing in our dataset yet connects a session signal to an exam result months later. Until it does, every statement we make about slide-level engagement is a statement about engagement, and stops there.

Attention time is a noisy proxy.

A slide left open is not a slide being read. The follow ratio is informative in aggregate and unreliable for any individual student, which is precisely why it never drives a per-student verdict in the product.

Single-institution data does not generalise.

Discipline, cohort size, room layout, language and local teaching culture all plausibly move these signals. This is the strongest argument for running the programme across an alliance rather than a department.

Observation changes the room.

Students who know their reactions are visible on a screen may behave differently for that reason alone. That is an interesting effect in its own right and a confound in everything else. Any serious design has to account for it.

European UnionEuropean Educational Technology

Built in Europe, aligned with theSustainable Development Goals

A live-classroom platform for innovative teaching, built in France and designed to adapt at large scale. Beyond desirability, feasibility and viability, Diaposi is measured against the UN 2030 agenda: impact is produced by the same mechanics that make the product work.

For a funding body this alignment is not decorative. These four goals are the frame most European instruments are assessed against, and in our case each maps to a mechanism already in the product rather than to an aspiration bolted on afterwards.

Sustainable Development Goal 4: Quality Education

Quality Education

The teacher finds out which concept failed during the lecture that failed it, not three weeks later in an exam.

Sustainable Development Goal 9: Industry, Innovation & Infrastructure

Industry, Innovation & Infrastructure

European research infrastructure for classroom interaction data, built to open standards rather than a private schema.

Sustainable Development Goal 10: Reduced Inequalities

Reduced Inequalities

No account, no install, any browser on any phone. Anonymity removes the social cost that silences the students who most need to ask.

Sustainable Development Goal 17: Partnerships for the Goals

Partnerships for the Goals

Ten institutions across ten countries already reachable through the INGENIUM alliance, which is the reason a multi-site study is realistic at all.

Partnership

Three ways in.

Different partners need different first steps. All three routes below are open right now, and the third one costs an hour.

Labs & doctoral schools

Host a funded doctoral thesis

The French CIFRE scheme funds a three-way contract between a company, a public laboratory and a doctoral candidate: €14,000 per year for three years, applications accepted year-round, roughly three months to a decision. We are actively looking for a co-supervising lab.

  • €42k over 3 years
  • Applications year-round
  • ~3 months to decision
Discuss a co-supervision

Universities & alliances

Run an instrumented cohort

Diaposi is already referenced across the INGENIUM European University Alliance: ten institutions in ten countries. That is a multi-site, multi-language research cohort that most groups spend two years assembling. Erasmus+ KA220-HED and the EIT Higher Education Initiative both fund exactly this shape of consortium.

  • 10 institutions, 10 countries
  • Erasmus+ KA220-HED
  • EIT HEI, up to €2m / project
Propose a consortium

Teachers

Start with one lecture

The lowest-barrier route into the programme, and the one we care about most. Run a single instrumented session in your own course, keep every row of the data, and co-author the case study. No account for your students, no institutional procurement, no commitment beyond the hour.

  • One session
  • You keep the data
  • Co-authored write-up
Run a session with us

Relevant instruments we are actively tracking: Erasmus+ Cooperation Partnerships (KA220-HED) and Alliances for Innovation, the EIT Higher Education Initiative, Horizon Europe Cluster 2, MSCA Doctoral Networks, and the French CIFRE scheme. If you are assembling a consortium for any of them, we would like to be in the room.

Data & governance

Your researchers get rows, not a PDF.

The fastest way to lose an academic partner is to hand them a dashboard screenshot. Session data exports as CSV and structured JSON today, one row per slide with every derived figure the interface shows.

Shipped

Raw export

Per-slide CSV and full session JSON, including attention times, votes, questions, polls and expectation ratings. No aggregation you cannot undo.

Committed

LTI 1.3 · xAPI · Caliper

LTI 1.3 / Advantage for launching inside Moodle and Canvas without an integration project. An xAPI event profile for research pipelines and Caliper Analytics 1.2 for institutional reporting.

Governance

DELICATE · SHEILA · GDPR

The Drachsler & Greller privacy checklist and the SHEILA institutional policy framework are the spine we expect to be assessed against, because that is what a university data protection officer will actually run us through.

References

What this page is built on.

The literature, frameworks and funding instruments cited above, in order of appearance in our own thinking rather than alphabetically.

  1. 01

    Chi, M. T. H., & Wylie, R. (2014). The ICAP framework: Linking cognitive engagement to active learning outcomes. Educational Psychologist, 49(4). Source

  2. 02

    Deslauriers, L., McCarty, L. S., Miller, K., Callaghan, K., & Kestin, G. (2019). Measuring actual learning versus feeling of learning in response to being actively engaged in the classroom. PNAS, 116(39), 19251–19257. Source

  3. 03

    Freeman, S., et al. (2014). Active learning increases student performance in science, engineering, and mathematics. PNAS, 111(23), 8410–8415. Meta-analysis of 225 studies: 1.5× higher failure rate under traditional lecturing. Source

  4. 04

    Nelson, C., et al. (2012). The effects of audience response systems on learning outcomes in health professions education. BEME Guide No. 21. Medical Teacher. Source

  5. 05

    Do audience response systems truly enhance learning and motivation in higher education? A systematic review (2025). Humanities and Social Sciences Communications. Source

  6. 06

    Castillo-Manzano, J. I., et al. (2016). Measuring the effect of ARS on academic performance: A global meta-analysis. Computers & Education. Source

  7. 07

    Drachsler, H., & Greller, W. (2016). Privacy and analytics: it's a DELICATE issue. A checklist for trusted learning analytics. LAK'16. Source

  8. 08

    Tsai, Y.-S., et al. (2018). The SHEILA framework: Informing institutional strategies and policy processes of learning analytics. Journal of Learning Analytics. Source

  9. 09

    A survey of multimodal learning analytics: Data, methods, systems, and responsible deployment (2026). Future Internet, 18(3), 115. Source

  10. 10

    From heartbeats to actions: Multimodal learning analytics of cognitive and behaviour engagement in real classrooms (2026). Learning and Instruction. Source

  11. 11

    Regulation (EU) 2024/1689 (AI Act). Annex III, high-risk: education and vocational training; Article 5, prohibition on emotion inference in education. Source

  12. 12

    1EdTech. Learning Tools Interoperability (LTI 1.3 / Advantage) and Caliper Analytics 1.2. Source

  13. 13

    SoLAR. LAK26, Bergen: Learning Analytics and AI Synergy. Source

  14. 14

    ANRT. Le dispositif CIFRE. Source

  15. 15

    European Commission. Erasmus+ 2026 call; EIT Higher Education Initiative 2026–2028; Horizon Europe Work Programme 2026–2027. Source