"Stand up straight! Chest out! Shoulders back!" These are the perennial cries of sergeant majors and fussy parents throughout the ages. Posture certainly matters. Big is dominant and in species after species, humans included, postures that enhance the posturer’s apparent size cause others to treat him as if he were more powerful. The stand up-straight brigade, however, often make a further claim: that posture affects the way the posturer treats himself, as well as how others treat him. To test the truth of this, Li. Huang and Adam Galinsky, at Northwestern University in Illinois, have compared posture’s effects on self-esteem with those of a more conventional ego-booster, management responsibility. In a paper just published in PsychoLogical Science they conclude, surprisingly, that posture may matter more. The two researchers’ experimental animals—77 undergraduate students—first filled out questionnaires, ostensibly to assess their leadership capacity. Half were then given feedback forms which indicated that, on the basis of the questionnaires, they were to be assigned to be managers in a forthcoming experiment. The other half were told they would be subordinates. While the participants waited for this feedback, they were asked to help with a marketing test on ergonomic chairs. This required them to sit in a computer chair in a specific posture for between three and five minutes. Half the participants sat in constricted postures, with their hands under their thighs, legs together or shoulders hunched. The other half sat in expansive postures with their legs spread wide or their arms reaching outward. In fact, neither of these tests was what it seemed. The questionnaires were irrelevant. Volunteers were assigned to be managers or subordinates at random. The test of posture had nothing to do with ergonomics. And, crucially, each version of the posture test included equal numbers of those who would become "managers" and "subordinates". Once the posture test was over the participants received their new statuses and the researchers measured their implicit sense of power by asking them to engage in a word-completion task. Participants were instructed to complete a number of fragments (for example, "l_ad") with the first word that came to mind. Seven of the fragments could be interpreted as words related to power ("power", "direct", "lead", "authority", "control", "command" and "rich"). For each of these that was filled out as a power word ("lead", say, instead of "load") the participant was secretly given a score of one point. Although previous studies suggested a mere title is enough to produce a detectable increase in an in dividual’s sense of power, Dr. Huang and Dr. Galinsky found no difference in the word- completion scores of those told they would be managers and those told they would be subordinates. The posture experiment, however, did make a difference. Those who had sat in an expansive pose, regardless of whether they thought of themselves as managers or subordinates, scored an average of 3.44. Those who had sat in constricted postures scored an average of 2.78. Having established the principle, Dr. Huang and Dr. Galinsky went on to test the effect of posture on other power-related decisions: whether to speak first in a debate, whether to leave the site of a plane crash to find help and whether to join a movement to free a prisoner who was wrongfully locked up. In all three cases those who had sat in expansive postures chose the active option (to speak first, to search for help, to fight for justice) more often than those who had sat crouched. The upshot, then, is that father (or the sergeant major) was right. Those who walk around with their heads held high not only get the respect of others, they seem also to respect themselves. The word "upshot" in the last paragraph most probably means ______.