OPINION3 January 2012
OPINION3 January 2012
Crawford Hollingworth on the observer effect and how people’s behaviour changes when they think or feel they are being watched.
There are certain social norms that we know are morally correct – such as paying for a train ticket or not littering – and not adhering to those rules could damage our reputation and our acceptance in society. Knowing we are being observed means that, by and large, we keep to these social norms and do the morally and socially responsible thing.
Incredibly, the act of observation can be simulated to produce similar effects. Good examples of this are the famous army recruitment posters from the First World War featuring an illustrated Lord Kitchener staring out at passers-by, informing them that “Your country needs you”. This series of posters first appeared in September 1914, and in that month thousands of volunteers signed up. It has inspired many imitations since then.
Roll on 90 years and researchers at Newcastle University decided to quantify the impact of this simulated observer effect. An experiment was conducted in their psychology department coffee room where people paid for their drinks by dropping money into an honesty box. Each week a new price list was published – but while the prices remained unchanged each poster carried a different picture, either of flowers or of faces looking directly at the drinker. All the pictures measured 15x3cm.
In weeks with eyes on the price list, staff paid on average 2.76 times as much for their drinks as in the weeks with flowers. “Frankly we were staggered by the size of the effect,” said Gilbert Roberts, one of the researchers involved in the study.
The images displayed each week are shown in the figure above. Looking at the graph, what is interesting is that the weeks when the eyes were clearest and most penetrating collected the most payments.
Such subliminal cues may have made people respond as if they were being watched and caused them to feel that they must be seen to be honest. Thus they were affected by social norms rather than feelings of altruism and fairness which honesty boxes are often thought to play on.
The same researchers have since found equally compelling evidence to show that including watchful eyes on canteen posters greatly improved rates of table clearing by diners (below).¹
Finally, a new study has shown how people who are lonely tend to conform with consumer norms if they feel others are observing their behaviour. “The Lonely Consumer: Loner or conformer?” by marketing professors Jing Wang (University of Iowa), Juliet Zhu (University of British Columbia) and Baba Shiv (Stanford University), published in the Journal of Consumer Research, found that lonely people are less influenced by what others have purchased than non-lonely people – but this depends on whether their purchases were made publicly or privately.
When buying in private, lonely consumers prefer minority-endorsed products as they fit better with feelings of loneliness. But when other shoppers could see their purchases, lonely people tended to conform to the majority and buy what others were buying.
“Lonely people’s preference for minority-endorsed products was only found when their preferences were kept private,” the authors write. “They switched to majority-endorsed products once their preferences became public.” The study reported that the research subjects tended to be concerned about being negatively evaluated by others.
These examples should prompt simple executional ideas for addressing social issues such as anti-littering campaigns or stopping graffiti and property damage in certain areas by using a broader understanding of social norms and how they can impact behaviour.
Crawford Hollingworth is a founder of The Behavioural Architects
1. Source: Bateson, M., Nettle, D.,Roberts, G., ‘Cues of being watched enhance cooperation in a real-world setting’, 2006, Biology Letters, doi:10.1098/rsbl.2006.0509,
Ernest-Jones, M., Nettle, D., Bateson, M. ‘Effects of eye images on everyday co-operative behaviour: a field experiment, Evolution and Human Behaviour 32 ( 2011 ) 172-178