← Back to hub
Hydrogen · Argyle Scott
The Science of Who’s Next

Seeing who can lead
through what comes next.

100 years of intelligence and personality research, condensed into a measurement system Hydrogen and Argyle Scott use to read leadership readiness across an entire workforce. With Psychometric Studio.

Hydrogen · Argyle Scott
Strand one , 1904

A puzzle about smart kids.

In the early 1900s, the psychologist Charles Spearman noticed something curious. When schoolchildren did well on a vocabulary test, they also tended to do well at arithmetic. And at spatial reasoning. And at memory. The correlations were everywhere, across tasks that seemed to have nothing in common.

He proposed a radical idea. A single underlying general intelligence , what he called “g” , was driving performance across all these different domains.

One invisible construct, measurable through many observable tasks, predictive across domains. This became the foundation of modern intelligence research.

Spearman (1904). General intelligence, objectively determined and measured. American Journal of Psychology.

Hydrogen · Argyle Scott
Strand two , 1988

A puzzle about rising executives.

Fast forward 80 years. Researchers at the Center for Creative Leadership spent a decade tracking promising executives as they moved up. Some thrived. Some derailed catastrophically.

The derailed ones often had the most impressive technical track records. The skills that had made them excellent managers , deep expertise, hands-on control, confident decision-making , became the very things that took them down at the next level.

The ones who succeeded shared one trait. They could let go of what had worked before, and learn what the new role demanded.

Lombardo & Eichinger (2000); McCall, Lombardo & Morrison, The Lessons of Experience (1988).

In practice: the senior leaders inside most organisations rose through deep technical or functional expertise. The pattern McCall and Lombardo identified is the central succession risk for any business promoting from within.
Hydrogen · Argyle Scott
Strand three , 2022

Two centuries.
One construct.

In 2022, Kenneth De Meuse published a paper in the Consulting Psychology Journal proposing something bold. That learning agility may function as the g-factor of leadership.

Just as Spearman’s “g” is the universal engine that drives cognitive performance across any thinking task, learning agility appears to be the universal engine that predicts leadership success across any unpredictable, complex, or first-time business environment.

For a long time this was an academic conclusion. Then the world caught up.

De Meuse, K. P. (2022). Learning agility: Could it become the g-factor of leadership? Consulting Psychology Journal, 74(3), 215-236.

Hydrogen · Argyle Scott
The world shifted

Academic theory became employer demand.

Every five years the World Economic Forum surveys over a thousand global employers , representing 14 million workers across 22 industries and 55 countries , and asks which skills will matter most.

WEF 2020 , Top Skills
1Analytical thinking and innovation
2Active learning and learning strategies
3Complex problem-solving
4Critical thinking and analysis
5Manual dexterity, endurance & precision
WEF 2025 , Fastest growing
1AI & big data
2Resilience, flexibility & agility
3Curiosity & lifelong learning
4Creative thinking
5Leadership & social influence
By 2025, employers report that 39% of core skills will change by 2030. The skills they identify as most differentiating growing roles from declining ones aren’t technical , they’re resilience, agility, curiosity, lifelong learning. Exactly the constructs De Meuse had spent two decades arguing were the substrate of leadership.

World Economic Forum, Future of Jobs Report 2020 & 2025.

Hydrogen · Argyle Scott
The question that’s shifted

The hiring question has
fundamentally changed.

It’s no longer “who has the skill today.”

It’s “who has the underlying capacity to learn the skill we don’t yet know we’ll need , and then learn the next one after that.”

That’s the question this framework was built to answer.

In practice: in any environment where the half-life of technical knowledge is shrinking, the cost of getting this question wrong compounds with every hire.
Hydrogen · Argyle Scott
The architecture

Leadership potential is
three things multiplied.

De Meuse identified three factors that together produce leadership potential. The critical insight: they’re multiplicative, not additive. Strength in one cannot fully compensate for absence in another.

Factor One
Cognitive capability
General mental ability, fluid intelligence, reasoning capacity. The raw horsepower. Necessary but never sufficient.
Factor Two
Learning agility
Willingness and ability to learn from experience, then apply those lessons in new situations. The engine of growth at every level.
Factor Three
Dark-side traits
Counter-productive patterns , arrogance, rigidity, emotional volatility , that surface under pressure and derail otherwise capable leaders.

Hired for intelligence. Promoted for performance. Derailed by dark-side traits. All three must be in view for a genuine read on potential.

Hydrogen · Argyle Scott
A finding worth pausing on

The self-awareness paradox.

Tasha Eurich’s large-scale study of self-awareness found that:

95%
of people think they are self-aware.
10-15%
actually are.

Which means most leaders are operating with a meaningfully distorted picture of how they are seen, how they make decisions, and where they get in their own way.

The interesting part? The areas where you most need self-awareness are, by definition, the areas you can’t yet see. That’s why reflection alone doesn’t fix it. Structured feedback does.

Eurich (2018). What Self-Awareness Really Is (and How to Cultivate It). Harvard Business Review.

Hydrogen · Argyle Scott
The framework

Nine facets. Three questions.

Dai & De Meuse (2021) organised learning agility into a clean 3×3 framework: three domains of learning × three components of agility. Every cell answers a specific question about a leader.

AbilityThe capacity
MotivationThe willingness
ApplicationThe adaptability
CognitiveHow you think
Mental Agility
Handles complexity, sees patterns, reasons flexibly
Change Agility
Seeks out novel and ambiguous situations
Results Agility
Delivers as conditions shift around them
SocialHow you connect
Social Astuteness
Reads political and emotional dynamics accurately
Open-Mindedness
Engages with different people, ideas, perspectives
People Agility
Works effectively across diverse styles and contexts
SelfHow you know yourself
Self-Awareness
Sees own patterns, reactions and impact accurately
Intellectual Curiosity
Explores ideas for their own sake , the engine of learning
Resilience & Composure
Sustains function under sustained pressure

Why this is diagnostic, not descriptive: read it across the rows. Capacity-without-deployment is a coaching problem. Willingness-without-capacity is a fit problem. Capacity and willingness without adaptability is a stress-tolerance problem. The grid tells you which conversation you’re in.

Dai & De Meuse (2021). Learning agility and the changing nature of leadership. Oxford University Press.

Hydrogen · Argyle Scott
A caveat that matters

Culture is the soil.

You can hire the most learning-agile people in the world. But if your culture punishes risk-taking, rewards consensus over candour, and treats failure as career death, then learning agility has nowhere to go.

Every facet has a cultural condition that either enables or suppresses it. Mental agility needs problems worth solving. Change agility needs real change to engage with. Resilience needs adversity that’s survivable. Self-awareness needs feedback that’s honest.

In practice: this is where the science meets the operating culture. Hiring the most agile people in the world doesn't return on that investment if the governance model punishes the kind of experimentation that lets agility show up.
Hydrogen · Argyle Scott
Why this matters legally, not just operationally

The compliance moment most
vendors haven’t priced in.

Two regulatory facts every European HR function now lives with.

GDPR Article 22
Restricts solely automated decisions
Where those decisions have legal or similarly significant effects on individuals. Hiring and promotion fall squarely inside this scope.
EU AI Act
Classifies HR & recruitment AI as “high-risk”
The second-highest category in the entire regulation. Documentation, audit-readiness, and human oversight requirements all apply.

Most assessment tools built in the last three years route candidate data through generative AI to produce their reports. That’s a regulatory exposure most vendors haven’t yet priced in.

Our architecture
Sidesteps it by design.
Every evaluative sentence in every report is human-authored from a narrative library. The engine selects; it never writes. Candidate data never enters any AI platform. If you’re ever audited on a hiring decision, you can produce the exact library, the exact weights, the exact derivation. That’s a defence. A black-box LLM report isn’t.
HR Hub · Senior HR Leaders
Why this matters for the data itself, not just the report

The integrity moment most
vendors haven’t built for.

In unproctored cognitive testing, roughly one in five candidates now show signs of attempted cheating. Self-reported AI use during pre-hire cognitive tests grew six-fold in the last twelve months. Whatever a vendor ships today, the next model solves. So the response can’t be a single defence, it has to be a layered architecture.

01 · Design
Dynamic item pools
Items rotate, so memorising or sharing them stops being an advantage. The test surface itself is moving.
02 · Deter
Pre-test integrity message
A single behavioural nudge at session start. In a field study of 3,072 matched candidates, flagged cheating behaviour dropped from 6.5% to 1.9%.
03 · Detect
Confidence Score
Behavioural signals during the session distinguish honest from AI-assisted candidates with 82% classification accuracy (Cohen’s d = 1.47).
04 · Defend
Facial validation
An identity check at the point of assessment, so the person being scored is the person being hired.
What this means for the AQ data
Behaviourally validated, not just assessed.
The new combined Podium & Psytech platform routes Adapt-g and the wider AQ assessment through this four-layer architecture. Each result arrives with a Confidence Score attached. If a candidate’s session looks behaviourally anomalous, the report flags it before it reaches the hiring conversation. Research claims drawn from Podium field studies (Shaban, Rose & Reid, SIOP 2026).
Hydrogen · Argyle Scott
From research to practice

Research without a tool
is still just research.

Everything we’ve shared with you today is research you can use regardless of whether you ever work with us. But after spending real time inside this body of work, we came to a simple conclusion.

The academic framework exists. The validated instruments exist. What was missing was the connective tissue , a transparent, deterministic way to turn trusted psychometrics into De Meuse’s nine facets, reliably, defensibly, and at the speed real hiring decisions move.

So we built it.

We call it the Agility Quotient: a psychometric assessment system that measures all nine facets, identifies derailment risks, and produces three different reports tailored to three different audiences and three different decisions.

1 / 13