認知風格對使用者介面設計之影響：以掃描器為例

The Influences of Cognitive Styles in User Interface Design: With Scanners as Examples

本研究的目的在於為不同認知風格之使用者建立合適之使用者介面設計準則。實驗樣本包括四廠牌的掃描器介面，其中包含了各類的介面型式，在資訊表達方面有：文字、圖像及文圖混用，在資訊結構方面有：頁籤與下拉式選單。受測者以CSA（Riding, 1991）測驗分為九類認知風格，並接受任務實驗，分別記錄這九類認知風格者對各任務的操作績效及主觀滿意度，並以MANOVA作為數據分析的統計方法。 結果顯示不同的認知風格者在使用每一個介面型式時的操作績效與滿意度方面均有顯著差異；在Wholist-Analytic認知風格向度上，Wholist的操作績效及滿意度以頁籤或下拉式選單較好，而頁籤則又比下拉式選單要更好一些；而Analytic的操作績效與滿意度則以主頁較好。至於在Verbal-Imagery認知風格向度上，Verbal的操作績效與滿意度依高低排列如下：文字、文圖混用及圖像；而Imager的操作績效與滿意度依高低排列則是：文圖混用、文字及圖像。

The purpose of this study is to construct proper set of criteria for interface design suitable for specific user groups. Four scanners chosen as experimental samples covered various interface designs, such as: texts, icons and mixed format in information representation, and tab and pull-down menu in information structure. Subjects were categorized into nine cognitive styles using CSA test （Riding, 1998）. A performance experiment and a questionnaire survey of subjective preferences were conducted to compare these interface designs among the nine cognitive styles. Finally, MANOVA was employed to analyze the data collected. The results indicate that a significant variation on both performance and preference in using each interface design was found among different cognitive styles. In the Wholist-Analytic dimension, Wholist users showed better performance and preference in using either tab or pull-down menu, and tab even better; while Analytic better in performing tasks without tab or pull-down menu. In the Verbal-Imager dimension, Verbal users showed descending priority order of text, mixed format and icon in both performance and preference measures; while Imager the order of mixed format, text and icon.