Imagine that you're taking a test in a large public hall. Obviously, your knowledge and confidence will determine your score, but could the number of people around you have an influence too? According to psychologists Stephen Garcia from the University of Michigan and Avishalom Tor from the University of Haifa, the answer is yes. They have found that our motivation to compete falls as the number of competitors rises, even if the chances of success are the same.
The simple act of comparing yourself against someone else can stoke the fires of competition. When there are just a few competitors around, making such comparisons is easy but they become more difficult when challengers are plentiful. As a result, the presence of extra contenders, far from spurring us on by adding extra challenge, can actually have the opposite effect. Garcia and Avishalom call this the "N-effect" and they demonstrated it through a number of experiments.
First, they showed that US students tended to score more highly in SAT tests in states where there were fewer people on average at each testing venue. For each state, they compared SAT scores in 2005 with the total number of test-takers divided by the number of venues, and adjusted the figures for factors such as education budget, general performance on the SATs and so on. A similar analysis of scores from the Cognitive Reflection Test (CRT) revealed the same pattern - a greater density of test-takers led to lower average scores.
Obviously, this is a very crude analysis. For a start, crowded testing venues could also be rife with distractions that could lie behind a dip in performance. Garcia and Avishalom knew that they had to come up with better evidence, so they ran an experiment.
They approached 74 students on their own, and asked them to complete a short quiz as quickly and accurately as possible. They were told that they were up against either 10 or 100 other students, and the top 20% would receive the princely sum of five dollars. Those who were pitched against a hypothetical 9 contenders completed the quiz in 29 seconds - significantly faster than the 33 seconds taken by those who were competing against 99.
This is clear example of the N-effect, of people behaving with different intents depending on how many others they thought they were competing against (none of whom were actually present). Garcia and Avishalom believe that the N-effect depends on people's propensity to compare themselves against their peers. The easier and more tangible those comparisons are, the more fuel there is for competition.
To demonstrate that, they told 50 students that they would have a week to win $100 by adding as many Facebook friends as possible. They found that the students felt more motivated to compete when facing 10 competitors compared to 10,000, and they were also more likely to compare themselves against the others within the smaller contest. The number of competitors predicted the students' motivations to compete, but that association disappeared after adjusting for their tendency to compare themselves with others.
This same experiment allowed them to rule out the possibility that the students were more motivated in the smaller group, simply because they thought the task would be easier. They certainly felt that way (albeit wrongly - in both cases, the prizes went to the top 20% and the students understood that) but it didn't affect their behaviour. Adjusting for this perception of difficulty didn't strongly affect the link between number of competitors and motivation.
Garcia and Avishalom admit that there are probably many other factors that lie behind the N-effect (and you may want to posit your own theories in the comments) but certainly, sizing yourself up against your peers is one of them. It's also unclear how far the effect extends. What are the smallest group sizes where the effect becomes apparent? When groups get larger, do you need larger differences to stimulate the effect? And does the effect apply across all forms of competition?
The answers to these questions will have to wait, but for now, Garcia and Avishalom suggest a couple of areas where the N-effect should be considered. In competitive workplaces (such as sales teams), workers may be more motivated and productive if they work in small branch offices rather than in one large, central location. In the education sector, the N-effect suggests that students may try harder in smaller classrooms (quite apart from the benefits of increased individual attention), and that they may even score slightly higher in exams if they are tested in smaller venues.
Reference: Garcia, S., & Tor, A. (2009). The -Effect: More Competitors, Less Competition Psychological Science DOI: 10.1111/j.1467-9280.2009.02385.x
- The peril of positive thinking - why positive messages hurt people with low self-esteem
- Our moral thermostat - why being good can give people license to misbehave
- To predict what will make you happy, ask a stranger rather than guessing yourself
- Attendance at religious services, but not religious devotion, predicts support for suicide attacks
- Lacking control drives false conclusions, conspiracy theories and superstitions
I've frequently noticed a strong random component on my grades, for individual exams and in overall courses. At least 50% of my undergrad & grad classes I feel my final grade greatly over- or underestimated my skill. Maybe at some level, individuals understand this and so approach competitive situations more as *lotteries* than a measure of mastery. (I think a similar random component always shows up in the few competitive sports I play as well. Sales also seem to be a strongly random vocation.)
If individuals do partially attribute their performance to luck, then one would expect them to try harder when the odds of "winning" are better, and just give up against poor odds. Many other explanations also fit the data at this stage :)
Great post Ed.
That's interesting, especially in the situations it may apply to. Gives me food for thought re my kids.
it seems like its not just competition, but also probability of winning. getting more facebook friends against 10 people, cool youd probably win. against 10,000, who cares you probably wont.
the SAT correlation was more interesting, as this test is supposed to reflect the individual's performance and capabilities more than a ranking/winning. Perhaps a talent contest would be more accurate to isolate one's competitiveness over perception of ranking probabilities (america's got talent?)
I enjoyed your post. I wonder if this N-effect has any correlation to the size of companies or the team sizes inside of companies? Under the assumption that smaller companies typically react to market changes faster than bigger ones, an individual may contribute more when they think they have a better chance at standing out amongst their peers inside a smaller company thereby adding to the competitive advantage of that company. So, in larger companies that align themselves with small nearly autonomous teams, they might be able to activate that N-effect to their advantage in same way that small companies inherently do. Google, IBM, and Proctor&Gamble come to mind as companies that consistently innovate and keep their teams small. They are all insanely profitable too. It might not be related to the N-effect, but I think if the research effectively proves the N-effect exists with test takers, then why not competition in the work place that has much more at stake than a couple bucks?
Does this mean that the smaller the class size the greater the competition? Hence, the pupils tend to work harder and get higher grades?
Is it possible that this phenonmenon is related to the same underlying cause of the bystander effect (people are less likely to help out in an emergency situation the more people that are around), social loafing, or even groupthink? That is there is a diffusion of "responsibility" - so that the larger the group the less individuals feel "responsible" for their behavior?