Does visual complexity (number of strokes, perimetric complexity, number of connected points, character type) matter in Chinese character recognition? : a mixed effect model analysis

Abstract

Visual analysis has a pronounced role on the early stages of Chinese character recognition and naming, yet the influence of visual complexity besides stroke number on Chinese characters remain unclear. The current study explored the effect of visual complexity in visual word recognition and naming of 4376 Chinese characters using four different complexity measures: number of strokes, perimetric complexity, number of connected points and character type. Mixed effect modelling was carried out to analyse the participants’ accuracy, latency and pupil dilation in a naming and lexical decision task. Result showed that the number of strokes, perimetric complexity and number of connected points are directly proportional to visual complexity. Low-frequency words were more affected by visual complexity where morecomplexed character with many stroke, high perimetric complexity, and many connected points, were recognized slower and more error prone than less-complexed characters. Also, each complexity measures was found to have their own distinct role facilitate the word recognition process. The current study revealed the importance to consider multiple visual complexity dimensions to provide a holistic overview of word recognition. Modification of traditional word recognition model is also needed to take into account the contributions of visual complexity.