The purpose of the present paper was to propose a practical method of revis
ing texts to enhance students' learning. The method used here was originate
d by Britton & Gulgoz (1991) and was translated to be more practical and ap
plicable to Japanese texts. It was applied to revise part of the junior hig
h school history texts so that the revised ones would be more locally coher
ent than the original ones. In the present experiment, either revised or or
iginal texts were assigned to 115 seventh grade students who had not learne
d its content. Their performance was measured in immediate and delayed cond
itions. The results showed that readers of the revised text performed bette
r than those of the original one. Educational implications from this study
were as follows: a) if students learned new things from texts, they would b
enefit more from ones with good local coherence (considerate texts) rather
than inconsiderate ones; b) the method used here to revise a text for local
coherence would be more practicable for an instructional text.