Proceedings of the Annual Meeting of the Cognitive Science Society

Previous studies have shown that when choosing one of thelogically equivalent frames (e.g., “half full” or “half empty”),speakers tend to choose one based on a reference point. Forexample, when the amount of water in a glass with 500ml ca-pacity was originally 0ml (or 500ml) and then increased (ordecreased) to 250ml, speakers tended to express the contentof water in the glass as “half full” (or “half empty”). We ex-amined why speakers chose one of the logically equivalentframes. In addition, we examined whether an irrelevant refer-ence point affected speakers’ choice of frame. In order to ex-amine these two issues, we conducted three behavioral exper-iments using a frame choice task. Specifically, participantswere presented with a task-relevant (story-based) or task-irrelevant (prime-based) reference point and then asked tochoose a frame. Following this, they were asked to reveal thereason for the frame choice. Our findings were summarizedwith the following two points. First, when reference pointswere task-relevant, many participants chose a frame based onthe reference point with explicit reason. Second, even whenreference points were task-irrelevant, they affected framechoices and almost all of our participants did not report the ef-fect of the irrelevant reference point. These results indicatethat the effect of reference points on frame choices is robustand that people do not always notice the effect.