閱讀差異分析信度與效度之探究與提升

The Validity and Reliability of Reading Miscue Analysis: Issues and Possible Consolidations

閱讀差異分析是Kenneth Goodman在1960年所提出的閱讀研究方法，可運用於閱讀歷程研究及閱讀理解評量。閱讀差異分析不同於大多數的標準化閱讀測驗，前者著重於閱讀理解的過程，而後者則偏向閱讀理解的結果。因差異分析具有信度與效度要求，故可視為研究閱讀過程和評量的重要工具。本研究旨在探討閱讀差異信度與效度之相關議題，故首先簡述閱讀差異分析為何，透過文獻評析與整合，討論其原有之信效度，並進而建議強化差異分析信效度的可能方式，以期幫助學者與教師善用差異分析進行閱讀研究與閱讀評量，使其成為標準化閱讀紙筆測驗之外的替代評量方法。

Miscue analysis is a reading research method and a reading evaluation tool developed by Kenneth Goodman in the 1960s. It has been used widely to study reading processes and evaluate reading performance. Generally considered an alternate reading assessment, miscue analysis reveals the processes of comprehending while most standardized reading tests show the results of comprehension. Although a valuable reading evaluation tool, miscue analysis provokes its share of concerns about test validity and reliability. The purpose of this study is to explore issues concerning the validity and reliability of miscue analysis research. It first presents a brief introduction to miscue analysis and procedures of collecting miscue data. Then issues concerning the validity and reliability of miscue analysis are discussed by surveying and synthesizing related research literature. This is followed by a discussion about possible ways to consolidate validity and reliability of miscue analysis. The study would help interested reading teachers and researchers to use miscue analysis as a counterpoint to other types of reading tests.