Mood Levels: The Science of Measuring Energy and Stress

How are you feeling?

It’s a simple question – but a powerful one. And at MOOD.ai, it’s where everything begins.

Because the truth is, most of us weren’t taught how to answer that question with nuance – or how to recognise our feelings as data worth noticing. That’s what we’re here to change.

When you check in with MOOD.ai, you’re not just picking a colour or an emoji – you’re tapping into a deeper emotional pattern that psychologists have been mapping for decades.

We believe emotions are essential. They don’t fit into neat boxes of good or bad – they exist on a spectrum. They're fluid, layered, and intelligent.


The Two Dimensions of Feeling

Psychologists have long searched for ways to describe the structure of emotional experience. One of the most widely accepted frameworks is the valence–arousal model, introduced by James Russell (1980). It maps how we feel using two core dimensions:

  • Valence: How pleasant or unpleasant the feeling is
  • Arousal: How activated or energised we feel

In the top right corner you’ll find emotions like joy (high valence, high arousal); in the opposite, emotions like sadness (low valence, low arousal). It’s not about categorising emotions as positive or negative – it’s about understanding where they sit in relation to each other, and how they shift over time.

This circular emotional space – known as the Circumplex Model of Affect – is now foundational in emotion research (Russell, 1980; Feldman Barrett & Russell, 1998).


Why We Measure Energy and Stress

At MOOD.ai, we reimagined how these axes show up in everyday emotional life.

While valence and arousal are useful in theory, they can feel abstract or clinical in practice. That’s why we use two terms that feel immediately familiar:

  • Energy – how activated, charged, or alert you feel
  • Stress – how tense, strained, or emotionally pressured you feel

This shift helps us reflect two key ingredients of mood: how much fuel you have, and how much pressure you’re under.

We’re not alone in this approach. Research by Schimmack and Grob (2000) found that arousal isn’t a single experience – it actually splits into two distinct types:

  • Energetic arousal – high energy with a positive tone (like excitement or motivation)
  • Tension arousal – high energy with a negative tone (like anxiety or frustration)

By treating Energy and Stress as separate but interacting dimensions, we’re able to capture a more accurate and human portrait of how people really feel – from bright yellow energy to deep blue depletion.


What's With the Colours and Emojis?

Each of the 12 zones in our mood model is represented by a unique colour and emoji pairing – and these weren’t chosen randomly.

They’re grounded in original research conducted at the University of Adelaide, where 1,589 participants linked colours and emojis with specific emotion concepts, alongside their own mental health scores.

This study revealed rich emotional patterns in how people visualise feelings. Yellow consistently reflected joy and energy. Blue evoked sadness and fatigue. Red captured stress and intensity. And softer tones like mint and cyan were associated with calmness, rest, and clarity.

These associations aren’t just aesthetically pleasing – they’re emotionally intuitive. They allow people to name and navigate what they’re feeling, even when words are hard to find.


Emotions Are Complex – and That’s the Point

MOOD.ai doesn’t try to flatten your feelings into a mood label. Instead, it invites you to check in with nuance – to notice, to reflect, and to understand.

Our dual-score approach using Energy and Stress helps you easily understand your daily mood levels, right from your dashboard.

Your mood isn’t good or bad – it’s just real. And when you begin to track how you feel over time, you start to see patterns.

With awareness comes insight. With insight comes choice. And with choice comes change. That’s the science – and the power – behind MOOD.ai’s approach to mood.

All emotions belong. Let’s keep feeling them.


References