Verguts Tom, De Moor Wendy
Department of Experimental Psychology, Ghent University, Belgium.
Exp Psychol. 2005;52(3):195-200. doi: 10.1027/1618-3169.52.3.195.
We investigate whether two-digit numbers are decomposed for purposes of numerical comparison (e.g., choosing the larger one). Earlier theorists concluded that numbers are processed holistically (Brysbaert, 1995; Dehaene, Dupoux, & Mehler, 1990), or that holistic and decomposed processes operate in parallel (Nuerk, Weger, & Willmes, 2001). In the present experiment, we presented pairs of two-digit numbers with a decade distance of either zero (e.g., 54-57) or one (54-61). If a holistic process contributes to two-digit comparison, there should be an overall distance effect for number pairs with a decade distance of one. On the other hand, if numbers are decomposed and a holistic comparison does not contribute, this overall distance effect should be absent for these number pairs. Evidence is found that, at least in the present task settings, numbers are not compared holistically. The results are interpreted in terms of a recently proposed theory of numerical cognition (Verguts & Fias, 2004; Verguts, Fias, & Stevens, in press).
我们研究两位数是否会为了进行数值比较（例如，选择较大的那个数）而被分解。早期的理论家得出结论，数字是被整体加工的（布里斯巴特，1995年；德阿纳、迪普克斯和梅勒，1990年），或者整体加工和分解加工是并行运作的（努尔克、韦格和威尔姆斯，2001年）。在本实验中，我们呈现了一对对两位数，其十位间距要么为零（例如，54 - 57），要么为一（54 - 61）。如果整体加工有助于两位数的比较，那么对于十位间距为一的数字对，应该会有一个总体的间距效应。另一方面，如果数字被分解且整体比较不起作用，那么对于这些数字对，这种总体的间距效应应该不存在。研究发现，至少在当前的任务设置中，数字不是被整体比较的。这些结果是根据最近提出的数值认知理论来解释的（韦尔古茨和菲亚斯，2004年；韦尔古茨、菲亚斯和史蒂文斯，即将出版）。
