本研究的主要目標是要探索整合式選擇題試題題型對學生在做局部理解試
題時所產生的影響。 實驗的方法之一是將整合式選擇題的文章打散，
使其題目成為單題選擇題試題題型。 整合式選擇題涵蓋整體理解及局部
理解試題兩類，局部理解試題並不會因失去文章的連貫性而受到影響，
在整合式選擇題及單題選擇題中均保有足夠的解題線索， 然而受試者卻
可能會因試題題型不同而有不一致的表現。 實驗的方法之二是對 於實驗
組的同學施以適當的教學， 說明整合式選擇題兩類試題的特性， 要求學
生注意解題線索， 再以整合式選擇題測量實驗組及控制組是否在局部理
解試題有不同的表現。 受試者是省立新竹女子高級中學一百七十位高三
學生。 根據實驗的結果， 我們得到以下的結論 :第一 : 整合式選
擇題試題題型會對受試者在局部理解試題解題上產生負面的影響。 受
試者在單題選擇題試題題型比在整合式選擇題試題題型的表現來的好且有
顯著差 異。 第二: 文體不會影響受試者在局部理解試題的表現。 第
三: 簡要的教導整合式選擇題的特性對局部理解試題解題有正面的影響
。 本研究的結果顯示適當的教學指導整合式選擇題試題題型是可以減輕
試題題型所產生的負面影響。

The present study aims to investigate whether the
multiple-choice (M-C)rational cloze test format affects EFL
learners' performance on non-globalitems. A total of 170
third-grade students at the Hsin-chu Girls' High Schoolserved as
subjects. Two experiments were conducted. In Experiment 1,
exposi-tory and narrative passages were selected and two
types of tests wereconstructed over each passage. In the
sequential type, items were presentedin the form of the M-C
rational cloze. In the scrambled type, the sentencesof each
passage were systematically shuffled and realized in the
discrete-point M-C test. Each cloze item appeared once in the
M-C rational cloze andonce in the discrete-point M-C test. All
subjects took both the discrete-point M-C test and the M-C
rational cloze. In Experiment 2, the experimentalgroup
received instruction in the techniques of solving non-global
items whilethe control group did not. Then, the M-C rational
cloze was administered toboth groups. Based on the results of
two experiments, the researcher came tothe following
conclusions : First, the M-C rational cloze test
formatnegatively afffects the subjects' performance on the
non-global items. Thesubjects comprehended the non-global
items in the discrete-point M-C testbetter than those in the
M-C rational cloze. Second, the text types have no-thing to
do with the statistically significant difference in the
subjects'poorer performance on the non-global items in the M-C
rational cloze than inthe discrete-point M-C test. Third,
the instruction in the techniques ofsomving non-global items
in the M-C rational cloze exerts influence on solvingthese
items in the narrative passage. Results of this study suggest
that theinstruction for the M-C rational cloze may be an
effective technique to reduceundesirable effect on non-global
items.
The present study aims to investigate whether the