中文字視知覺測驗之發展與心理計量特性驗證

Abstract

前言：書寫是學齡兒童日常生活中的一項重要任務，而視知覺則是成功完成書寫的基本要件。若視知覺能力不足，可能對書寫表現產生負面影響，進而損害學業成就及心理健康。儘管視知覺的重要性已受到重視，然而，目前的工具仍無法全面且精確地評估中文書寫所需的視知覺能力。因此，本研究旨在開發一個專為評估兒童中文書寫相關視知覺能力的測驗工具。此外，本研究亦檢驗此工具的心理計量特性，包括內部一致性、再測信度、測量誤差及建構效度，以確保其在臨床應用中的可靠性與有效性，進而促進視知覺相關問題的早期診斷與介入。 方法：中文字視知覺測驗（Chinese Characters Visual Perception Test, CCVPT）包含三個分測驗：視覺辨識、視覺空間知覺、視覺記憶。題目設計基於中文字的視覺特徵，包括階層結構與組字原則。為評估該測驗之心理計量特性，本研究招募小一與小二學生共51名。其中20位學生於兩週內接受再測，以檢驗再測信度與最小可偵測變化值 (minimal detectable change, MDC)；另有23位小一學生於六個月後再次接受評估，以探討其對未來書寫表現的預測能力。建構效度藉由測試五個假設進行檢驗。 結果：中文字視知覺測驗在所有分測驗 (Cronbach’s α = 0.74–0.80) 及總分(Cronbach’s α = 0.89) 皆展現良好的內在一致性。在低年級學童中，再測信度結果顯示視覺辨識（ICC = 0.62）與視覺記憶（ICC = 0.61）分測驗具中度信度，視覺空間知覺分測驗（ICC = 0.77）與總分（ICC = 0.87）則具良好信度。測量誤差在各分測驗中屬可接受範圍 (MDC95% = 13.43%–16.76%)，而總測驗的測量誤差則達優異水準 (MDC95% = 8.56%)。結果支持中文字視知覺測驗的建構效度，五項預設假設中有四項（80%）被證實。本測驗：(1) 與既有視知覺測驗（視知覺技巧測驗第四版）之間具低至中度相關；(2) 與視覺動作整合測驗（拜瑞－布坦尼卡視覺動作整合發展測驗第四版）之間具低至中度相關；(3) 與書寫表現呈現低至中度相關，包括書寫速度（以基本讀寫字測驗抄寫測驗中的字數計算）、正確性（以基本讀寫字測驗抄寫測驗中正確字數的百分比計算）、易讀性（兒童中文書寫易讀性測驗）及家長評估的書寫表現（兒童寫字表現評量表）；以及 (4) 與書寫表現的相關性高於視知覺技巧測驗第四版。然而，對於 (5) 預測書寫表現隨時間變化的能力，支持證據有限。 結論：中文字視知覺測驗融合中文字的視覺特徵，於國小低年級學童中展現良好的信度與效度，顯示其在評估與中文書寫相關之視知覺能力具備潛力。整體而言，本測驗具有良好的臨床應用價值，適用於學齡初期兒童之視知覺功能評估。建議未來研究可進一步探討其於不同年齡層的適用性，以驗證並拓展本研究之發現。

Introduction： Handwriting is an essential occupation for school-aged children, with visual perception playing a crucial role in its successful execution. Deficits in visual perception abilities can negatively impact handwriting proficiency, potentially affecting academic achievements and mental health. Despite the importance, the existing visual perception assessment tools might be insufficient to address the specific demands of Chinese handwriting. Thus, the purpose of this study was to develop a new assessment tool specifically designed to evaluate the visual perception abilities related to Chinese handwriting performance in children. Furthermore, the psychometric properties of the tool, including internal consistency, test-retest reliability, measurement error, and construct validity, were examined to ensure its clinical applicability in facilitating the early diagnosis and intervention of visual perception-related issues. Methods： The Chinese Characters Visual Perception Test (CCVPT) comprises three subtests: visual identification, visual-spatial perception, visual memory. The items were designed based on the features of Chinese characters, including their hierarchical structures and compositional principles. To evaluate its psychometric properties, 51 first- and second-grade students were recruited. Among them, 20 students were assessed twice within a two-week interval to estimate test-retest reliability and minimal detectable change (MDC). Additionally, 23 first-grade students were reassessed after a six-month interval to investigate its predictability for handwriting performance. Construct validity was examined through testing five hypotheses. Results： The CCVPT showed good internal consistency across all subtests (Cronbach’s α = 0.74–0.80) and the total score (Cronbach’s α = 0.89). The coefficients of test-retest reliability were moderate for visual identification (ICC = 0.62) and memory subtests (ICC = 0.61), and good for visual-spatial perception subtest (ICC = 0.77) and total score (ICC = 0.87). Measurement error was acceptable for all subtests (MDC95% = 13.43%–16.76%) and excellent for the total score (MDC95% = 8.56%). Construct validity of the CCVPT was supported, with four out of five (80%) predefined hypotheses confirmed. Specifically, the CCVPT demonstrated: (1) weak to moderate correlations with existing measures of visual perception (Test of Visual Perceptual Skills–4th Edition, TVPS-4); (2) weak to moderate correlations with visual-motor integration (Berry-Buktenica Developmental Test of Visual-Motor Integration–4th Edition); (3) weak to moderate correlations with handwriting performance, including speed (measured by word count in the copy subtest of the Battery of Chinese Basic Literacy, BCBL), accuracy (percentage of correct words in the BCBL), legibility (Chinese Handwriting Legibility Assessment for Children), and parent-reported performance (Chinese Handwriting Evaluation Form); and (4) stronger associations with handwriting performance compared to the TVPS-4. However, limited support was found for the 5th hypothesis regarding the CCVPT’s ability to predict handwriting performance over time. Conclusions： The CCVPT, which incorporates the visual features of Chinese characters, demonstrated good reliability and validity among lower-grade elementary school children. These findings suggest its potential as a tool for assessing children's visual perceptual abilities, particularly in relation to Chinese handwriting performance. The results support the CCVPT’s clinical utility and its suitable for evaluating visual perception in early school-aged children. Future studies are recommended to examine its applicability across different age groups to support and extend the current findings.