Evidence, not vibes.
Usage is not effectiveness. A teacher can complete every session and change nothing in their classroom. This page sets out, honestly, what AIM Education measures, the validated instruments behind those measurements, and where the evidence currently sits on a public tier ladder.
Guskey's five levels: each one depends on the last
The evaluation spine is Guskey's five levels of professional-development evaluation. Each level is a necessary link. Evidence at a lower level does not guarantee the next.
| Lvl | What it measures | What AIM Education captures |
|---|---|---|
| 1 | Participant reaction | TAM-derived per-unit micro-surveys (relevance, credibility, usefulness). |
| 2 | Participant learning | Baseline and per-unit knowledge checks, compared between the start and end checks of each unit. |
| 3 | Organisation support & change | Adapted EdTech Context Inventory items on leader expectations and follow-up. |
| 4 | Use of new knowledge | Self-reported integration items (Woldemariam et al. 2025) + stored teacher artefacts. |
| 5 | Student outcomes | Aspirational, pursued only with a research partner and full consent. |
The 5Es, impact dimensions
The 5Es describe what "impact" means in full. All five are adopted here, weighted for a teacher-PD product.
| Dimension | Guiding question | Where it stands now | Where it is going |
|---|---|---|---|
| Efficacy | Does it work in controlled conditions? | Out of near-term scope. | Year 3+ aspiration. |
| Effectiveness | Does it work in the real world? | Pedagogical fit, feasibility (Bronze). | Correlational pre/post (Silver). |
| Ethics | Is it safe, ethical, transparent? | Plain-language data practices, granular consent, no third-party tracking. | External ethics review. |
| Equity | Who benefits and who is left out? | Built to WCAG 2.1 AA; cross-context relevance. | Subgroup analysis as cohorts grow. |
| Environment | Does the school context support it? | Leader instruments capture conditions at entry + exit. | School-level case studies. |
The evidence scorecard
A public tier ladder, Bronze, Silver, Gold, set against the EdTech evaluation literature. It says plainly which tier I am working towards and what would be needed to climb.
Documented theory of change, programme design grounded in published research, measurement items adapted from validated instruments, and a platform built to WCAG 2.1 AA.
Correlational pre/post evidence of teacher knowledge and practice change, across the first cohorts.
Independent, controlled study, a multi-year, partnership-dependent aspiration. I will not claim it before I can evidence it.
Measurement, link by link
Every measurement fires on a progress milestone, never on a calendar date. Thirteen units, self-paced online, with optional live online sessions.
Each unit has a start check and an end check, alongside programme anchors. The deliverable check is factual and never attitudinal, do you now have X, yes, partly or no. It never asks whether something was useful, and that is the reason the numbers next to it can be trusted.
1. Baseline (before Unit 1)
Knowledge, confidence, and school context, the anchor every later measurement is compared against.
2. Start and end checks
Two short checks bookend each unit. Triggered by progress milestones, never by calendar dates.
3. Artefacts
Teacher-generated outputs (guardrails, tool evaluations, AI policies, implementation plans) stored as first-class evidence.
4. End of programme
Impact on practice and an honest written reaction. No satisfaction score is used as a headline.
5. 6-week follow-up
What actually stuck, re-asks the integration items 6 weeks on. Practice change shows up after the programme ends.
Instruments this is built on
- EdTech Align (Fegely et al. 2023), teacher competency, used for L2 ethics and digital-literacy items.
- EdTech Context Inventory (Kohler et al. 2022), school conditions, used for L3 leader instruments.
- Woldemariam et al. (2025), integration-frequency and readiness items, used for L4 self-reported practice change.
- TAM-derived perceived-usefulness items, used for L1 unit reaction surveys.