本研究首先整合閱讀理論與傳統文本分級研究，建立一「文本適讀性分級架構」，此架構 包含印刷、語言與文學、體裁、內容及多樣性等五大構面，以及其下的十七類準則及五十八 個適讀性指標。接著以層級分析法探討25 位閱讀領域的研究與教學專家對架構中各構面、準 則與適讀性指標之於文本適讀性重要性的認知。結果顯示：「內容」為影響文章理解最重要的 構面，其次依序為「語言與文學」、「體裁」、「多樣性」與「印刷」構面。整體權重值最高的 前五個適讀性指標為「真實性」、觀點「一致性」、「觀點數」、概念「深度」及「大眾文化」； 最低為「文本尺寸」。本研究建立的「文本適讀性分級架構」，包含影響理解的各個重要構面， 並可以客觀量化適讀性指標的重要性，提供更周延的文本分級。

In this study, reading theories and traditional text leveling systems were integrated and a text leveling framework comprising five aspects (i.e., print features, language and literary features, text structure, content, and diversity), 17 criteria, and 58 indicators was developed. Subsequently, 25 experts in the reading research and reading education fields were invited to assess the importance of these aspects, criteria, and indicators to text leveling. Data were analyzed using the analytic hierarchy process. The results show that content was the most critical aspect influencing text comprehensibility, followed by language and literary features, text structure, diversity, and print features. Overall, the five indicators in this framework with the highest global weight were truthfulness, consistency of perspective, number of perspectives, depth of concepts, and popular culture. Text length exhibited the lowest global weight. The text leveling framework established in this study includes all the major dimensions affecting comprehension and, through objectively quantifying the importance of leveling indicators, provides highly comprehensive text leveling.