Wedell Douglas H, Fitting Sylvia, Allen Gary L
Department of Psychology, University of South Carolina, Columbia, South Carolina 29208, USA.
Psychon Bull Rev. 2007 Aug;14(4):681-6. doi: 10.3758/bf03196821.
The participants were 40 students who were briefly presented 32 dot locations, one at a time, and attempted to reproduce each location after a short delay. Half of the participants completed the task with the surrounding shapes being a circle, a horizontal ellipse, and a vertical ellipse; for the other half, the surrounding shapes were a square, a triangle, and a pentagon. Elongation of the task field along an axis led to exaggerated bias along that axis, but the pattern of bias was fairly constant across the shapes. The data were modeled by assuming that bias in estimation was due to the weighting of spatial category prototypes. Modeling indicated that shape affected spacing of prototypes, but there was no evidence that it affected the number of prototypes. These results were consistent with use of a viewer-based frame of reference, with prototypes reflecting four spatial quadrants generated by left-right and up-down distinctions from the viewer's perspective.
参与者为40名学生，他们被短暂呈现32个点的位置，每次一个，并在短暂延迟后尝试重现每个位置。一半参与者在周围形状为圆形、水平椭圆和垂直椭圆的情况下完成任务；另一半参与者的周围形状则为正方形、三角形和五边形。任务区域沿某一轴的伸长会导致沿该轴的偏差增大，但偏差模式在不同形状之间相当一致。通过假设估计偏差是由于空间类别原型的加权来对数据进行建模。建模表明形状会影响原型的间距，但没有证据表明它会影响原型的数量。这些结果与基于观察者的参照系的使用一致，原型反映了从观察者角度由左右和上下区分产生的四个空间象限。
