數字技術及其伴隨的人工智能分析技術正在改變學習評價方式。數字化評價更具實施效率,更容易推廣,能更有針對性地反映個體層面的表現,更貼近學習環境,更具互動性,并且支持更具想象力的、豐富多彩的、互動性強的、及時的反饋。不過,本文認為"更多、更快、更美"只是21世紀初期學習評價的一部分。學校亟需在學習內容和評價內容方面做出相應轉變,而學生則需要完成更復雜的學習任務,這些學習任務超越認知領域,超出對知識內容的掌握,更強調對能力的塑造和對技能的掌握,包括軟技能或通識能力等。本文試圖探討這一學習評價的前沿領域,研究能否以及如何利用學習管理系統及其他數字學習工具所生成的大量數字化、以過程為導向的數據,對學習者掌握復雜能力和通識能力的程度進行可靠且有效的判斷。本文認為,確定開發學習評價工具的測量分析標準是確保學習評價工具之信效度的關鍵。

Digital technology and artificial intelligence(AI)analytical technologies are changing the assessment of learning.Digitized assessment can be more administratively efficient, more broadly implemented, more reflective of individual levels of performance, more integrated into the learning environment,more interactive,and more supportive of timely feedback.However,this paper argues that the abovementioned assessment type reflects only the first quarter of the 21 st century.Educational institutions are pressed to make distinctive shifts in what is learned and what is assessed.Students need to establish the ability of complex learning that extends to individual capacity and skills,including soft skills or knowledge in the liberal arts.This paper explores this assessment frontier and examines how large quantities of digital data generated from learning management systems and other learning tools can be used to make valid judgments about the degree to which student have mastered complex knowledge.It argues that‘metrolytic’standards for development of assessment tools can be applied to ensure such validity.