The effect of emotions on your behaviour depends partly on your expectations

Will you get a better result from a business negotiation if you get angry or remain calm? What about a creative task – will you come up with more solutions to a problem if you're excited, or relaxed?

The answer, according a new study in Emotion, is that it depends at least in part on what you expect the impacts of emotions to be.

Some theories linking emotion and behaviour hold that emotions activate fixed behavioural "programmes" (anger activates aggressive actions, for example). Others hold that while emotions do influence behaviour, how they do so depends upon the individual's past experiences, and the current context. (Faced with a bullying boss, the anger you feel may lead you respond aggressively, if this has worked for you in the past; alternatively, it may prompt you to go off and strengthen bonds with colleagues.)

Maya Tamir and Yochanan Bigman at the Hebrew University in Jerusalem, Israel, reasoned that if this second class of theory is correct, a given emotion will lead to a particular behaviour if a person expects it to do so – but if that expectation is not there, it won't.

For the first experiment, the researchers told volunteers that, as part of a study on peer influence, they would be completing a real negotiation task, with financial consequences. The negotiation was over the sharing of coloured chips. Each participant was given a table showing the monetary value of the chips, but they were also told that the other person's table may not show the same values.

Before going into the negotiation, half were assigned to an "angry" condition. They read testimonials purporting to be from people who had previously been successful at the negotiation task that included comments like, "Throughout the negotiation, I was persistent. Eventually I got angry and my partner felt compelled to give me what I wanted." They also listened to "angry" music (music rated in a pilot study as inducing feelings of anger), such as Inquisition by the cello metal – yes, cello metal – band Apocalyptica. The control group read testimonials that did not mention emotions, but recommended being "reasonable". They listened to music deemed to be emotionally neutral. Finally, the volunteers described their current emotional state, then went into the negotiation.

The researchers found that the volunteers who said they felt angry made more money, but only if they'd read the testimonials advocating the benefits of anger, not if they hadn't. Those who'd read the emotionally neutral testimonials did equally well, whether they felt angry or not.

In another experiment, Tamir and Bigman recruited 159 more students, to look at how expectations about anger might affect performance on a first-person shooter video game.  Half the group was led to believe, via accounts from previous players, that feeling angry would help them kill more baddies, while the other half was led to believe that feeling calm would be beneficial. Those who expected anger to boost their performance played better if they actually felt angry, rather than relatively calm; in contrast, angry participants in the "calm is good" condition didn't show this advantage. However, against the researchers' expectations, feeling calm didn't help the 'calm is good' group perhaps because in this context, the lower physiological arousal associated with calmness is a disadvantage, regardless of beliefs.

For the final study, the researchers switched to creativity. Among people who were encouraged to believe that excitement benefits creativity, those who reported feeling excited did better (as measured by how many uses they could come up with for a brick) than those who reported feeling calm. In contrast, among the "calm is good" group, those who felt calm performed better than those who felt more excited – showing that there are contexts where a low arousal state can be advantageous if you believe it to be so.

Clearly, more work is needed to explore just what types of performance can be influenced by expectations about the impacts of emotions, and actual emotional states. But this work does show that, as the researchers conclude, "At least in some cases, what we expect emotions to do may determine what they actually do."