本研究主要的目的在以多重方法探討高中生閱讀兩難情境故事的因果推論
歷程，此種歷程包含線上與離線兩部分。本研究以建構論與最小量論對閱
讀中自動化推論的屬於整體或區域諧和性議題為基礎，探討多個可能前因
的文章表面距離影響因素。本研究以156名高中生為研究對象，分別接受
口語原案，原因決定測驗或實驗等多重方法的施測。其主要方法與結果有
下述二項：1、促發實驗部分採用3(目標項性質)*2(文章呈現方式)的混合
實驗設計，並以ANOVA進行考驗。其結果顯示：閱讀中的文章表徵狀況較
符合建構論的整體諧和性預測；2、原因決定測驗部分，以卡方進行考驗
。結果顯示：閱讀本研究的兩難故事後，受試較傾向選擇近結果的原因；
回憶狀況的此種選擇傾向，會隨線上文章呈現方式的不同，而略有不同的
選擇簡單易取原因傾向。最後，研究者針對研究結果進行討論，並提出實
用上以及未來研究上的建議。

The purpose of this paper was to investigate processes of causal
inferencesthat senior high school students school students made
in reading dilemma narratives. The processes included two parts,
on-line and off-line. This study investigated inferences of
distant factors for antecedent causalalternatives in texts,
based on different concepts of automatic global andlocal
inference during texts reading between constructionist and
minimalist. The subjects of the study were 156 senior high
school students. They were tested by multimethods, namely,
verbal protocal, cause-decision test, and priming experiment.
The main methods and results of the study were as follows:1. A
3(Target Items)* 2(Ways of Text Presentation) mixed design was
adopted in priming experiment, and ANOVA was used to analyze the
data. The results were consistent with the global-coherence
prediction of constructionist. 2. Chi-square statistic was used
to test the results of the cause-decision test. After reading
the dilemma narratives, subjects tended to choose more adjacent
causal antecedents. And in recalling, they tended to make the
easy-available choices, varying with ways of text presentation.
Discussions and suggestiond were offered at the end of this
paper.
The purpose of this paper was to investigate processes of causal