Triplett noticed in 1898 that cyclists rode faster when racing against others than when racing against the clock alone. The finding seemed straightforward: competition motivates. What Zajonc’s (1965) social facilitation theory established, sixty-seven years later, was that competition was not the active ingredient. Mere presence was. Even non-competitive audiences, people simply watching without any competitive dynamic, changed what performers produced. And not always in the same direction. The relationship between observation and performance turned out to be more conditional, and more commercially consequential, than the simple observation that people try harder when watched.

Zajonc’s mechanism: arousal, task type, and the direction of the effect

Zajonc’s account of social facilitation resolved what had been a contradictory literature. Some studies found that audiences improved performance. Others found they degraded it. Zajonc’s insight was that both findings were correct, and the variable that determined the direction was task type.

Audience presence produces arousal. Arousal facilitates the production of dominant responses, the responses that are most practised, most automatic, and most available. On well-practised tasks with a clear correct response, arousal improves performance: the skilled typist types faster, the experienced presenter delivers more fluently, the trained athlete executes more cleanly. On novel, complex tasks where the correct response is uncertain and exploratory thinking is required, arousal impairs performance by strengthening the dominant response at the expense of the flexible, generative thinking the task demands. The same mechanism that helps the expert hurts the learner. The same observation that sharpens execution degrades exploration.

Cottrell’s revision: the critical variable is evaluation, not presence

Cottrell, Wack, Sekerak and Rittle (1968) refined the account by identifying the specific feature of audiences that produces the effect. Mere presence of non-evaluative observers produced weaker facilitation effects than the presence of clearly evaluating audiences. Blindfolded observers, audiences who were evidently paying no attention, produced attenuated effects. The active variable was evaluation apprehension: the belief that the audience was assessing the quality of the performance and forming a judgment about the performer.

This revision has direct implications for every observation-heavy professional environment. The remote worker whose keystrokes are being logged by productivity monitoring software is not merely being watched. They are being evaluated. The evaluation apprehension that Cottrell identified activates the same stress response that Dickerson and Kemeny’s (2004) research on social-evaluative threat documents: cortisol rises, the prefrontal cortex goes partially offline, and the cognitive flexibility required for complex work narrows precisely when the observation is most intense. The monitoring that was introduced to improve productivity is impairing the most valuable work the employee does.

The Hawthorne effect: observation as the intervention

The Hawthorne Works studies produced the most commercially cited observation finding, though the mechanism has been refined considerably since Mayo’s (1933) original account. Workers who knew they were being observed changed their behaviour regardless of what specifically was being changed in their working conditions. The awareness of observation was itself sufficient to alter performance.

The commercial application is more specific than the popular account suggests. Observation does not uniformly improve performance. What it reliably does is make the observed person aware of being watched, which activates evaluation apprehension when the observation is perceived as evaluative and activates the self-consciousness mechanism when the observation is salient. Either pathway changes what is produced. The manager who introduces a performance dashboard and sees productivity figures improve has not necessarily discovered that their team was previously underperforming. They may have discovered that visible measurement, even without any attached consequence, changes behaviour through the observation mechanism alone.

Self-consciousness and the disruption of automaticity

Duval and Wicklund’s (1972) self-awareness theory identifies the specific mechanism through which audience observation impairs expert performance on well-practised complex skills. Audience presence activates self-monitoring: attention shifts from the task outcome toward the process of executing the skill. For genuinely expert performers, this shift is destructive. Expert performance on complex skills depends on automaticity, the ability to execute the skill without conscious monitoring of each component step. The expert golfer who thinks about the mechanics of their swing while swinging is no longer executing an automatic skill. They are consciously managing a process that functions best outside conscious management. The audience has done the opposite of what was intended.

The sports psychology literature on choking documents this at scale. The most technically accomplished players under the most scrutinised conditions, penalty kicks in World Cup finals, critical putts in major championships, show performance degradation that correlates precisely with the degree of evaluative pressure and the extent to which attention has been directed toward process rather than outcome. The skill that was practised until it was automatic becomes unreliable the moment observation makes it deliberate again.

Observation and insight: why creativity requires privacy

The Kounios and Beeman (2009) insight neuroscience research identifies what is perhaps the sharpest practical implication of the observation-performance relationship for knowledge workers. The neurological state that precedes genuine insight involves alpha wave activity that actively suppresses external sensory input: the brain closes off external stimulation to allow the internal associative processing that produces the sudden connection between previously unrelated ideas. The condition required for insight is the opposite of the condition produced by observation.

Observation activates external attention. Insight requires its suppression. The open-plan office that creates continuous low-level observation, the camera-on video meeting that makes every participant’s face and reactions visible to every other participant, and the productivity monitoring system that makes every action trackable all create precisely the external attention activation that the insight research predicts will prevent the internal associative processing that generates the most valuable cognitive work. Research on camera-on versus camera-off video meetings consistently documents lower creative contribution rates in camera-on conditions, which the evaluation apprehension and insight-suppression mechanisms together predict.

The task-type framework and its managerial implications

The Zajonc-Cottrell framework produces a practical structure for thinking about when observation helps and when it hurts. Observation of simple, well-practised, objectively measurable tasks tends to improve performance through the arousal and evaluation apprehension mechanisms: the data entry operator who knows their speed and accuracy is being measured will likely perform better on that specific measure. Observation of complex, novel, creative, or strategic tasks tends to impair performance through the same mechanisms, because the tasks require the exploratory thinking and psychological safety that evaluation apprehension suppresses.

This means that the manager who applies observation uniformly across their team’s work, monitoring all activity with equivalent intensity, is simultaneously improving performance on the tasks that matter least and degrading performance on the tasks that matter most. The structural implication is not no observation, but differentiated observation: high transparency and measurement for routine execution, and genuine protected privacy for the complex cognitive work that requires the conditions observation destroys.

Books worth reading on this

Choke by Sian Beilock is the most comprehensive available account of what happens to performance under pressure and observation, covering the specific cognitive and neurological mechanisms through which evaluation apprehension, self-consciousness, and working memory disruption produce the performance failures that the social facilitation and choking literatures document. Beilock is a cognitive scientist who has spent her career studying performance breakdown in high-stakes contexts, from elite athletes to university students in high-pressure examinations, and her account moves between the laboratory research and the real-world performance contexts where the mechanisms operate with unusual clarity and specificity. Her treatment of the automaticity disruption mechanism, and the specific conditions under which directing attention toward process rather than outcome degrades expert performance, is the most detailed and accessible available account of what Duval and Wicklund’s self-awareness theory predicts at the level of actual performance in actual high-stakes situations. For the entrepreneur or team leader who wants to understand not just that observation changes performance but the precise conditions under which it helps versus hurts, and what can be done at the individual and organisational level to manage the mechanisms involved, Beilock provides the most practically grounded available treatment of a literature that is usually confined to sports psychology and cognitive science journals.

If the dynamics described here are significantly affecting your wellbeing, speaking with a psychologist is the right next step. UK: Samaritans (116 123, free, 24/7). Mind (0300 123 3393). BACP: bacp.co.uk/search/Therapists. Crisis Text Line — text HOME to 741741 (US, UK, Canada, Ireland). International: internationaltherapistdirectory.com.

This article is for educational and informational purposes only. Sources: Triplett, N. (1898), The Dynamogenic Factors in Pacemaking and Competition, American Journal of Psychology, 9(4), 507-533. Zajonc, R.B. (1965), Social Facilitation, Science, 149(3681), 269-274. Cottrell, N.B. et al. (1968), Social Facilitation of Dominant Responses by the Presence of an Audience and the Mere Presence of Others, Journal of Personality and Social Psychology, 9(3), 245-250. Dickerson, S.S. & Kemeny, M.E. (2004), Acute Stressors and Cortisol Responses, Psychological Bulletin, 130(3), 355-391. Mayo, E. (1933), The Human Problems of an Industrial Civilisation, Macmillan. Duval, S. & Wicklund, R.A. (1972), A Theory of Objective Self-Awareness, Academic Press. Kounios, J. & Beeman, M. (2009), The Aha! Moment, Current Directions in Psychological Science, 18(4), 210-216. Beilock, S. (2010), Choke, Free Press. Csikszentmihalyi, M. (1990), Flow: The Psychology of Optimal Experience, Harper & Row.