本研究旨在編製「臺灣兒童中文詞彙深度測驗」，使其得以使用於母語為中文的小學一至六年級的兒童，以補充國內測量詞彙深度工具的缺乏。測驗編製的理論基礎來自 Schmitt (2014)的詞彙深度工具的理論。本研究分為預式與正式階段，在預式階段，受試者為國小一至四年級的學生，人數共有60人，預式題目共有40 組試題(240個子題)；在正式階段，測驗試題共有32組試題，受試者總計242位國小一至六年級的兒童，進行團體施測。本研究以古典測驗理論 (CTT)及試題反應理論(IRT)進行試題分析，並統計學生的能力值及平均答對率。試題分析以及 Rasch 模式分析指出，本測驗的試題難度偏易，具有良好的鑑別度，且測驗試題與 Rasch 模式的適配程度良好。在信度方面，內部一致性Cronbach′s α 係數為 .94、PCM信度為.93。在效度方面，與畢保德圖畫詞彙測驗之同時效度為.52，同時與其月齡之相關為.51。綜合上述，本測驗具有良好之信度及效度，並顯示出本測驗可有效測量國小一至六年級學生的詞彙深度。此測驗將有助於未來兒童的詞彙、閱讀發展及教學診斷、介入等研究。;This study aims to develop the Taiwanese Children′s Chinese Depth of Vocabulary Test of elementary school students in Taiwan. The test is designed to address the lack of vocabulary depth assessment tools for Chinese-speaking students in grades one to six. The theoretical framework for test development is based on Schmitt′s (2014) Depth of Vocabulary Tool. The study consists of two phases: a pilot phase and a formal phase. In the pilot phase, 60 students from grades one to four participated, and a set of 40 test items (240 sub-items) was administered. The formal phase involved 242 elementary school students from grades one to six, and a total of 32 test item sets were administered in a group setting. The test items were analyzed using Classical Test Theory (CTT) and Item Response Theory (IRT). Descriptive statistics were used to analyze the students′ ability values and average item response rates. The item analysis and Rasch model analysis indicated that the test items were relatively easy in difficulty, had good discriminability, and showed a good fit with the Rasch model. The reliability analysis yielded an internal consistency Cronbach′s α coefficient of .94 and a Person Separation Index reliability of .93. In terms of validity, the test demonstrated concurrent validity with the Peabody Picture Vocabulary Test (r = .52) and a correlation with age (r = .51). Overall, the Taiwanese Children′s Chinese Depth of Vocabulary Test exhibited good reliability and validity, and it effectively measured depth of vocabulary of elementary school students in grades one to six. This test will be valuable for future research on children′s vocabulary and reading development, as well as educational diagnosis and intervention.
