Alexander Thomas and Stella Chess conducted one of the most important longitudinal studies in developmental psychology — the New York Longitudinal Study, begun in 1956 — which followed 133 children from infancy through adulthood and produced a nine-dimension framework for describing temperament. Their work established that infants arrive with measurable, stable individual differences in how they respond to the world, and that these differences shape — but don't determine — later development. The nine dimensions they identified remain the most detailed and empirically grounded description of infant and early childhood temperament available.
The Origins of the Thomas and Chess Framework
Thomas and Chess began their longitudinal study in reaction to what they saw as an excessive emphasis on environmental determinism in the psychoanalytic and behavioural traditions dominant in the 1950s. Both movements tended to treat the child as a relatively passive recipient of environmental influence. Thomas and Chess hypothesised that children are active participants in their own development, and that their constitutionally given temperament shapes how they interact with caregivers and environments — and therefore what environments they end up creating for themselves.
Their method involved structured parent interviews and direct behavioural observations, coding behaviour across nine dimensions from the earliest months of life. The nine dimensions were derived inductively from the data rather than imposed from a prior theory — which gave them unusual empirical grounding compared to more theoretically derived frameworks.
The Nine Temperament Dimensions
1. Activity Level
The degree of physical activity, motility, and movement in a child's behaviour across waking states. High-activity children are physically busy, explore through movement, and find sedentary demands challenging. Low-activity children are quieter and more settled. This dimension persists into adult personality as energy level and restlessness — the adult who can't stay still at a desk or the one who is comfortable with stillness for extended periods.
2. Rhythmicity (Regularity)
The predictability of biological functions — sleep cycles, hunger, bowel function. Rhythmic children are highly regular in their biological schedules; arrhythmic children are unpredictable. This dimension matters practically for caregiving — a highly rhythmic infant is significantly easier to schedule. It also relates to adult capacity for routine and the ability to maintain consistent habits.
3. Approach/Withdrawal
The nature of the initial response to new stimuli — people, foods, situations, places. Approach children move toward novelty readily; withdrawal children pull back initially. This is perhaps the most significant dimension for predicting adult social behaviour and is closely related to Jerome Kagan's concept of behavioural inhibition, which is one of the most extensively studied temperament variables. Initial withdrawal doesn't prevent eventual engagement — it describes the initial response, not the stable end-state.
4. Adaptability
How easily a child adjusts after the initial response. A child who initially withdraws from new food may adapt quickly and accept it within a few exposures (high adaptability) or may reject it consistently over many exposures (low adaptability). This dimension is distinct from approach/withdrawal — a child can have an initial positive approach but adapt slowly to sustained demands, or withdraw initially but adapt quickly once the novelty passes.
5. Intensity of Reaction
The energy level of responses — whether positive or negative. High-intensity children are dramatically expressive in both pleasure and distress. Low-intensity children's reactions are muted regardless of emotional valence. This dimension shapes how visible the child's states are to caregivers, which in turn affects the quality of care they receive — a highly intense, clearly signalling infant may receive more responsive care than a low-intensity infant whose distress is less legible.
6. Threshold of Responsiveness
How much stimulation is needed to produce a response — the sensitivity threshold. A low-threshold child responds to subtle stimuli that high-threshold children don't register. This dimension relates to adult sensory sensitivity, emotional reactivity, and the capacity for noticing environmental details. Elaine Aron's highly sensitive person concept is closely related to this dimension.
7. Quality of Mood
The characteristic positive or negative quality of a child's mood across situations. This is baseline affective tone — not reaction to specific events but the underlying emotional weather. Positive-mood children are generally cheerful, smiling, and pleasant; negative-mood children are more frequently fussy, serious, or crying. This dimension is closely related to adult neuroticism and hedonic baseline.
8. Distractibility
The degree to which external stimuli interfere with ongoing activity. Highly distractible children are easily pulled off task by any new stimulus. Low-distractibility children maintain focus despite peripheral activity. This dimension has obvious relevance for learning, task completion, and attention regulation — and for understanding children who struggle in stimulating environments.
9. Attention Span and Persistence
The length of time a child pursues a task and the degree to which they continue in the face of obstacles. Highly persistent children keep working at something even when it's frustrating. Low-persistence children move on quickly. This dimension is closely related to adult conscientiousness, grit research, and the psychological study of self-regulation.
Thomas and Chess's Three Temperament Clusters
From their data, Thomas and Chess identified three broad clusters that grouped many of the children, while acknowledging that roughly 35% of children didn't fit neatly into any cluster:
- Easy children (approximately 40% of their sample) — positive mood, high rhythmicity, initial approach to novelty, high adaptability, low-to-moderate intensity. These children were the easiest to parent and adjusted well to most environments.
- Difficult children (approximately 10%) — negative mood, irregular rhythmicity, initial withdrawal, slow adaptability, high intensity. These children placed the greatest demands on caregivers and were more likely to develop behavioural problems — though the outcome depended heavily on fit with the caregiving environment.
- Slow-to-warm-up children (approximately 15%) — initially withdrawn, slow to adapt, but of low intensity and eventually positive mood once comfortable. These children didn't need the same active management as difficult children but required patient, gradual introduction to new situations.
Goodness of Fit: The Central Contribution
Thomas and Chess's most influential conceptual contribution was the goodness-of-fit model: the idea that outcomes for children depend not on temperament alone but on how well the demands and expectations of the environment match the child's temperamental characteristics. A "difficult" child is not destined for poor outcomes — but they do require caregiving that matches their high intensity and slow adaptability. A "slow to warm up" child in a patient, low-pressure environment thrives; the same child under pressure for rapid adjustment does poorly. The framework shifted emphasis from temperament as destiny to temperament as personality characteristics requiring appropriate matching.
To understand your own personality dimensions that trace back to these early temperament qualities, our free Big Five personality test gives a full profile across the dimensions most thoroughly studied in adult personality research.
Frequently Asked Questions
Are the nine dimensions still used in research today?
Yes, though typically in modified form. The nine-dimension framework has been influential in developmental paediatrics and child psychiatry. Subsequent research has sometimes reduced the nine dimensions to fewer, more robustly distinct factors — particularly using three higher-order dimensions (negative emotionality, surgency/extraversion, and effortful control), which map onto combinations of Thomas and Chess's original nine.
Can temperament change over development?
Thomas and Chess found substantial stability in temperament traits across development, but not perfect stability. The dimensions most stable across time are activity level, initial approach/withdrawal (behavioural inhibition), and quality of mood. More context-sensitive dimensions like distractibility and rhythmicity show more variability. The general consensus in developmental psychology is that temperament is moderately stable — more stable than most environmental influences would predict, less stable than fully fixed biological traits.
Is "difficult temperament" a disorder?
No. Thomas and Chess were careful to distinguish temperament characteristics from disorder diagnoses. The "difficult" cluster describes a pattern of traits that create challenges for caregiving — not a pathological condition in the child. Whether a difficult-temperament child develops behavioural problems depends heavily on the quality of the caregiving environment. Many children with this profile become highly capable, intense adults in appropriate contexts.
How are Thomas and Chess's nine dimensions related to ADHD?
The high-distractibility, low-persistence, high-activity, and high-intensity dimensions in Thomas and Chess's framework overlap substantially with ADHD symptom descriptions. Thomas and Chess themselves discussed the relationship, arguing that what becomes diagnosable ADHD is often a severe expression of temperament traits that exist on a continuum, combined with environmental demands that amplify the challenges these traits create. Their goodness-of-fit model suggests that many ADHD-related difficulties are at least partly environmental — about mismatch between temperament and educational or caregiving expectations.
Do Thomas and Chess's dimensions predict adult personality traits?
Broadly yes, with moderate correlations. The New York Longitudinal Study follow-up data showed significant continuities between infant/child temperament measures and adult personality. Activity level in infancy correlates with adult energy and extraversion. Negative mood quality correlates with adult neuroticism. Approach/withdrawal correlates with adult introversion/extraversion and anxiety. The correlations are not perfect — adult character development significantly shapes how temperament expresses itself — but the temperamental substrate remains visible.