Abstract
We do not yet know why adults do not always take others’ perspectives automatically. If adults use their own perspectives as an anchor, they should show similar egocentrism for both desire and knowledge perspective taking tasks. To explore this previously untested possibility, I tested paradigms for measuring perspective taking for desires in adults, and investigated factors that might affect the chance of egocentrism.
I first asked participants to estimate the percentage of peers who would know about or like a song, expecting participants to estimate that more peers would like or know about the song when the participant did than when they did not (consensus effects). This hypothesis was supported. However, participants tended to underestimate the actual percentage of participants in the sample liking or knowing about the song, perhaps due to motivation to be part of a distinct social group.
Later, I investigated whether high collectivism and low individualism, or primed interdependent self construal, might reduce egocentrism. Increasing vertical individualism was indeed associated with increasing expected consensus. However, priming did not appear to affect consensus estimates. Because the independence prime might have primed a difference mindset instead, I primed participants to think about similarities or differences, predicting that participants primed to think about differences should show smaller consensus effects than the others. No such effect was found.
I also asked participants to make gift-giving decisions in three separate tasks. In one, participants predicted how much an individual would like each of a set of gifts, which varied in how clearly they were suited to the recipients’ interests. Recipient-participant similarity was also varied. I expected participants to predict greater appreciation for gifts they liked than gifts they disliked, and that this effect would be stronger for gifts with ambiguous suitability than for those with clearer suitability, and would increase with increasing recipient-participant similarity. As predicted, participants predicted greater appreciation for gifts they liked than gifts they disliked. Contrary to expectations, no effects of ambiguity, and no strong effects of similarity, were found on participants’ egocentrism.
In a second gift-giving task, participants chose between two potential gifts for each recipient. The recipient’s preferred gift (the target) was ethically controversial, whereas the less preferred gift (the alternative) was not. Ethical opposition to the target was expected to increase the chances of choosing the alternative. There was a non-significant trend in this direction. Increasing personal responsibility for the choice was also expected to increase such an effect. Instead, only relative price had any consistent effect on choices.
In a third gift-giving task, participants selected a colour for a gift from a continuum. Recipients’ colour preferences were presented implicitly. Participants’ choices were expected to be biased towards their own preferred colour. This did not happen, but participants took longer to choose when they did not share the recipient’s preferences.
I concluded that models that feature an initial reliance on one’s own perspective, such as anchoring and adjustment models, cannot be ruled out as explanations of adult egocentrism, but the egocentric effects of one’s own desires are probably weak.