Data Methodology - ProgramFinder

1. Overview

ProgramFinder scores Ontario college programs on a 1–100 scale using a five-pillar AI resilience model. The score reflects how well-positioned a program's graduates are to remain employable as AI and automation tools mature through the late 2020s.

Scores are intended as a comparative research aid — a single, structured signal to help students prioritise and shortlist programs. They are not a career guarantee, an admissions rating, or an indicator of program quality.

Neutral by design

No institution pays to influence a score. The methodology is applied identically to all 194 programs.

Fully transparent

Every pillar, weight, and data source is documented here. You can reconstruct any score from first principles.

Periodically updated

Scores are reviewed when significant labour market shifts or new NOC data are published.

For the student

Designed to be understood in 30 seconds. A higher score means a more durable career path — full stop.

2. The Five Scoring Pillars

Each program is rated 1–10 on five independent factors. The composite score is a weighted sum scaled to 100. Each factor captures a dimension of automation resistance that is supported by labour economics research.

01

Human Interaction

25% weight

Measures how much the role requires face-to-face human contact, empathy, relationship management, or real-time social judgment.

Why this matters: AI systems remain poor substitutes for genuine human connection. Roles with high human contact — healthcare, counselling, early childhood education — have consistently resisted automation pressure even as back-office tasks in those fields have been automated.

Rating scale

1–3 Minimal human contact — primarily digital or solo work

4–6 Moderate contact — regular client or colleague interaction

7–8 High contact — sustained empathetic or relational work

9–10 Core human factor — the role is almost entirely about human relationships

Examples: Personal Support Worker (9), Social Service Worker (8), Computer Programming (3)

02

Physical & Hands-On Work

20% weight

Assesses the degree to which the role requires physical presence, manual dexterity, spatial reasoning, or working with physical materials, equipment, or environments.

Why this matters: Robotics and physical automation remain expensive to deploy outside controlled, repetitive environments. Trades roles in unstructured settings — construction sites, dental offices, aircraft hangars — are structurally difficult to automate even when the software side of the field is highly disrupted.

Rating scale

1–3 Entirely desk-based or digital

4–6 Regular physical tasks alongside office or digital work

7–8 Primarily physical — tools, equipment, or spatial environments

9–10 Highly physical — safety-critical manual precision or physical labour

Examples: Dental Assisting (9), Welding and Fabrication (9), Accounting (2)

03

Licensing & Regulatory Protection

20% weight

Rates the degree to which practice in the field is governed by statutory licensing bodies, regulated health profession acts, professional certifications, or safety-critical regulatory frameworks.

Why this matters: Licensed professions create a legal barrier to entry that AI-generated output cannot cross — a machine cannot hold a licence, carry malpractice insurance, or be professionally accountable under Ontario's Regulated Health Professions Act. Heavily licensed programs produce graduates who are structurally protected from direct AI displacement of their role.

Rating scale

1–3 No licensing requirement — open practice

4–6 Voluntary certification or soft professional standards

7–8 Mandatory provincial or federal licensing

9–10 Strictly regulated — statutory body, scope-of-practice limits, mandatory insurance

Examples: Respiratory Therapy (9), Practical Nursing (9), Graphic Design (2)

04

AI Adaptability

20% weight

Measures how well the program's graduates can leverage AI as a productivity tool rather than being replaced by it — including digital literacy, adaptable skill sets, and ability to work alongside AI systems.

Why this matters: Programs whose graduates are comfortable integrating AI tools into their practice will see productivity gains rather than displacement. This pillar rewards programs that develop transferable, conceptual, and supervisory skills alongside technical execution — the human-in-the-loop advantage.

Rating scale

1–3 Role is directly replaceable by current AI tools

4–6 Partial AI integration possible; some tasks automated

7–8 Graduate likely to use AI as a tool to increase output

9–10 Strong supervisory or creative role — graduate directs AI systems

Examples: Advanced Care Paramedic (8), Police Foundations (7), Data Entry (2)

05

Routine Digital Exposure

15% weight

Rates the degree to which the role primarily involves processing structured digital information through predictable, rule-based digital workflows — the category most susceptible to AI automation.

Why this matters: Large language models, document processing AI, and robotic process automation are already displacing roles whose primary function is reading, classifying, or transforming structured digital content. This pillar acts as a penalty factor — a high score here reduces the overall AI resilience score.

Rating scale

1–3 Minimal routine digital work — primarily physical or human-facing

4–6 Regular data entry, reporting, or digital admin tasks

7–8 Primary role is digital information processing

9–10 Almost entirely routine digital — high automation target

Note: This pillar is inverted — a higher rating reduces the overall score.

Examples: Respiratory Therapy (3 — good), Data Entry (9 — high risk), Nursing Unit Clerk (4)

3. Composite Score Formula

AI Resilience Score (0–100)

25 × Human
Interaction + 20 × Physical &
Hands-On + 20 × Licensing &
Regulation + 20 × AI
Adaptability − 15 × Routine
Digital

Each pillar is rated 1–10. Weights sum to 100. The Routine Digital Exposure pillar is subtracted (penalty factor). Maximum achievable score: 100. Minimum: 1.

Worked example: Respiratory Therapy

Pillar	Rating (1–10)	Weight	Contribution
Human Interaction	9	×25	+22.5
Physical & Hands-On	7	×20	+14.0
Licensing & Regulation	9	×20	+18.0
AI Adaptability	8	×20	+16.0
Routine Digital Exposure	3	×15	−4.5
Composite score (raw ÷ 10 × 100)			80 / 100

Why these weights? Human Interaction leads at 25% because the labour economics literature most consistently identifies empathy and relationship work as the hardest category to automate. Licensing and Physical work each take 20% because they create structural barriers independent of skill level. AI Adaptability at 20% recognises that automation augments some workers even while displacing others. Routine Digital Exposure is capped at 15% as a penalty to avoid over-penalising hybrid roles.

4. Career Outlook Data

Source

Career outlook ratings are sourced directly from Employment and Social Development Canada (ESDC) through the Job Bank Employment Outlook for the Ottawa–Gatineau Census Metropolitan Area, covering the 2025–2027 projection period.

Employment and Social Development Canada (ESDC)
Job Bank Employment Outlook · Ottawa–Gatineau CMA · 2025–2027
jobbank.gc.ca ↗

Outlook star ratings

ESDC assigns a 1–5 star employment outlook to each NOC occupation code. We map these directly to the tool's star display:

Stars	ESDC Label	What it means
★★★★★	Very Good	Significant job openings expected, employment growth above average.
★★★★	Good	Steady job openings, employment growth at or above average.
★★★	Moderate	Balanced supply and demand; some openings expected.
★★	Limited	Employment growth below average; candidate surplus likely.
★	Very Limited	Weak demand; significant candidate surplus expected.

NOC mapping

Each college program is manually mapped to the most relevant NOC 2021 occupation code. Where a program prepares graduates for multiple NOC codes, the primary placement occupation is used. NOC codes are reviewed annually.

Important limitations

ESDC outlook data covers the Ottawa–Gatineau region only. Conditions in Toronto, Hamilton, or London may differ materially.
Outlook projections are based on labour force surveys and econometric modelling. They are projections, not guarantees.
Some programs do not map cleanly to a single NOC code. These are assigned 0 stars and labelled "Undetermined".

5. Median Wage Estimates

Median wage figures displayed in the Career Outcome Analysis modal are category-level estimates derived from ESDC Job Bank wage data and Statistics Canada Labour Force Survey data for the Ottawa region.

Health $72,000 Regulated health professions, PSW, paramedics

Technology $74,000 Software, IT, networking, data

Trades $68,000 Electrical, welding, HVAC, aircraft maintenance

Public Safety $65,000 Police foundations, corrections, emergency management

Business $55,000 Accounting, marketing, HR, administration

Media & Design $50,000 Graphic design, animation, photography, broadcasting

Community Services $48,000 Social work, addictions, early childhood education

Hospitality & Culinary $42,000 Culinary management, hotel management, events

Sport & Recreation $44,000 Fitness, sport management, outdoor recreation

Environment $56,000 Environmental studies, forestry, geomatics

These are estimates, not guarantees. Actual starting salaries vary significantly by employer, experience, location, union status, and negotiation. Verify current salary data with the college admissions office, Ontario college career services, and Statistics Canada.

6. What Scores Don't Measure

The AI resilience score intentionally focuses on a specific question: how automation-resistant is a graduate of this program likely to be? As a result, it does not capture:

Program quality or reputation

Scores are independent of college rankings, student satisfaction, or graduate employment rates reported by the college.

Admission difficulty or selectivity

A high score does not indicate a harder program or higher admission standard.

Starting salary accuracy

Category-level wage estimates are not a substitute for program-specific salary data from the college.

Personal suitability

A score of 90 does not mean the program is right for you. Work style, aptitude, and passion are beyond the model's scope.

National or international conditions

Scores reflect Ottawa-region employment outlook data. Other Canadian cities or international job markets may differ.

Post-2027 AI developments

The model is calibrated to current AI capabilities. Rapid changes in AI — positive or negative — may shift scores significantly.

Part-time or self-employment markets

NOC data measures salaried employment. Freelance or entrepreneurial paths in a field are not captured.

7. Update & Review Process

New ESDC NOC data

Career outlook ratings are updated when ESDC publishes a new Employment Outlook cycle (typically every 1–2 years).

Material AI capability shifts

If a specific AI tool demonstrably changes the automation risk profile of a category of programs, the affected pillar ratings are reviewed.

New Ontario programs

When Ontario colleges launch new programs in our tracked categories, they are added and scored using the same methodology.

Feedback and corrections

If a scoring error is identified — for example, a program that was incorrectly mapped to a NOC code — it is corrected and the version number incremented.

Every update increments the dataset version number shown in the tool header. Historical scoring versions are archived and available on request.

8. Independence Statement

No institution pays to influence scores.

ProgramFinder operates independently of all Ontario colleges and post-secondary institutions. College participation in the Growth or Premium partner programme unlocks lead routing and analytics features — it does not affect, adjust, or influence program scores in any way.

The five-pillar scoring methodology is applied identically to all 194 programs regardless of whether the college has a commercial relationship with us, whether the college is aware of the tool, or whether the college has requested changes to their scores.

If you believe a score is factually incorrect — for example, because a licensing requirement has changed — please contact us. Score corrections are welcomed and acknowledged. Contact us →