兒童中文書寫易讀性評估之發展

背景與目的：書寫易讀性為書寫文字易於他人閱讀的程度。書寫易讀性不佳，可能阻礙書寫的表達或得到較低的評分。根據文獻回顧，中文書寫易讀性評估包含問卷與電腦化評估兩種形式，問卷可能受填答者的主觀判斷影響，而電腦化評估目前尚無法分析字體內部部件間的書寫表現。因此需發展中文書寫評估工具，以補足過往評估的限制。本研究目的共有二項：(一) 發展兒童中文書寫易讀性評估，以紙本的形式，客觀地評估孩童的中文書寫易讀性表現。(二) 驗證兒童中文書寫易讀性評估工具的心理計量特性。方法：共150位國小低年級孩童參與本研究。本研究分為兩個階段：(一) 發展階段：經由文獻回顧並採納專家意見，將書寫易讀性的評估分為五個面向：「大小」、「位置」、「旋轉」、「部件大小」及「部件空間」，並透過另外計分的筆畫面向作為評估的參考。(二) 心理計量特性驗證：施測者內信度、施測者間信度、再測信度、因素效度、收斂效度，以及區辨效度。結果：兒童中文書寫易讀性評估具有良好的施測者內信度及施測者間信度 (ICC > 0.80)。總分及筆畫的再測信度良好 (ICC = 0.83, 0.74)；五個分向度的再測信度尚可至良好 (ICC = 0.52 ~ 0.82)。因素效度的結果，「旋轉」面向的因素負荷量過低 (λ= 0.20)，其餘四個面向皆具備可接受的標準 (λ= 0.48 ~ 0.53)，而將「旋轉」面向刪除。收斂效度的結果，與兒童寫字表現評量表具低至中度顯著相關。區辨效度結果顯示，可區辨一般發展孩童與書寫困難孩童之差異。討論：兒童中文書寫易讀性評估，具備良好的信度及效度，可透過紙本的形式，客觀地評估孩童之書寫易讀性表現，以了解孩童書寫易讀性的問題類型與嚴重程度，作為後續書寫易讀性相關介入內容之參考。

關鍵字

中文字；書寫；易讀性；評估工具；心理計量特性

並列摘要

Background and purpose: Handwriting legibility represents the readability of the handwriting products. Problems in handwriting legibility may affect a child’s communication in written formation and receive poorer grades from teachers. Chinese legibility assessments include questionnaire- and computer-based tools. However, information obtained from questionnaire is limited to subjective perception and computer-based test were unable to analysis radials within the Chinese character. Therefore, this study aims to develop the Chinese Handwriting Legibility Assessment for Children (CHLAC) and to examine the psychometric properties of the CHLAC. Methods: 150 elementary school students in grade 1-2 participated in this study. This study consisted of two phase. (1) Development phase: 5 domains were developed for scoring the legibility, including size, position, orientation, radical proportion, and radical space. Besides, an additional domain, stroke, gives supplementary information of handwriting legibility. (2) Examination of psychometric properties phase. Results: The CHLAC had good intra-rater and inter-rater reliabilities (ICC > 0.8). Test-retest reliabilities in total score and the stroke domain were good (ICC > 0.7), while adequate to good in the 5 domains (ICC = 0.52 ~ 0.82). Construct validity showed the factor loading of orientation domain was low (λ = 0.20), but the other 4 domains were acceptable (λ = 0.48 ~ 0.53). Therefore, the orientation domain was deleted from the CHLAC. The convergent validities showed low to moderate correlations with the CHEF. The CHLAC presented good discriminant validity to differentiate children with and without handwriting legibility problems (p < 0.001). Conclusion: The CHLAC presented good reliabilities and validities. The CHLAC may provide objective measurements of handwriting legibility through paper-based assessment that might provide guidelines for handwriting legibility intervention.