The Effects of Reading Speed on Visual Search Task

The Effects of Reading Speed on Visual Search Task Masaharu Kato (pieko@abmes.twmu.ac.jp) Tokyo Women’s Medical University 8-1, Kawada-cho, Shinjuku-ku, Tokyo162-8666, JAPAN Mitsugu Kuriyama (kuri@cs.c.u-tokyo.ac.jp) Kazuhiro Ueda (ueda@gregorio.c.u-tokyo.ac.jp) The University of Tokyo 3-8-1, Komaba, Meguro-ku, Tokyo 153-8902, JAPAN Toyofumi Sasaki (sasaki@sokudoku.co.jp) NBS Japan Society of Speed Reading Education 3-6-2, Shibuya-ku, Tokyo150-0002, JAPAN Hirokazu Atsumori (h-atsu@rd.hitachi.co.jp) Hideo Kawaguchi (kawaguti@rd.hitachi.co.jp) Advanced Research Laboratory, Hitachi, Ltd. 2520 Akanuma, Hatoyama, Saitama 350-0395, JAPAN Yukuo Konishi (ykonishi@abmes.twmu.ac.jp) Tokyo Women’s Medical University 8-1, Kawada-cho, Shinjuku-ku, Tokyo162-8666, JAPAN Abstract Speed reading training enhances one’s reading speed. Typically, those who train using the Park-Sasaki training method can read all the characters sequentially with the reading speed of over 10,000 characters per minute (about over 4,000 words per minute in English sentences). It means that they take in a huge amount of information during speed reading. Our study investigated the training effect on visual perception and attention using a visual search task and revealed that speed reading changes the performance of a visual search task. In the conjunction task, compared with novices, the slope of response times for experts as a function of set size was shallow, which suggests that experts watch a wider area than do novices, or that experts conduct a parallel search, at least in part. The measurements of eye movements during the experiments support this explanation; experts could respond with fewer episodes of fixation than could novices. As we used a simple non-letter-based feature as the stimulus, the results suggest that training improves general ability to process visual stimuli, regardless of language. We think that speed reading training improves attention which plays an important role in the conjunction task. Introduction Speed reading training enhances one’s reading speed. Typically, those who train using the Park-Sasaki training method can read all the characters sequentially with the reading speed of over 10,000 characters per minute (about over 4,000 words per minute in English sentences). The trained people assert that they understand the contents of the books read at such high speed. If true, besides comprehending sentences, it means that they take in a huge amount of visual information during speed reading. That is interesting in terms of training and visual attention. To take in a huge amount of visual information, they must have a larger spotlight of attention than normal readers have, and/or they must move their spotlight of attention more quickly than normal readers do. As trained people acquire their speed reading ability by training, not by talent, this attentional change might be brought by the training. However, no study has examined the effects of speed reading training on visual attention [for a review, Carver The aim of this study is to verify that trained people can really take in visual information faster than normal readers can, and to investigate whether they have a larger spotlight of attention or they move it more quickly than normal readers, by using visual search task. The visual search task was introduced by Treisman and Gelade (1980), and has become a widely used measure in the study of visual perception and attention. In a speeded visual search task, subjects can look at objects (called items) on a screen for as long as they like, and a target that differs from uniform distractors in only one feature dimension is presented on the screen. The response time usually does not increase with the number of distractors. This is usually regarded as a reflection of spatially parallel processing (Treisman & Gelade, 1980). Treisman and Gelade (1980) argued that if a target is similar to a distractor (e.g., conjunction), the response time is largely dependent on the number of distractors, which they called serial processing. Although subsequent research has shown that a conjunction search can be quite efficient in some cases (e.g., Nakayama and