This is an empirical study to examine the efficacy of giving learners information on "error perception and interpretation," which is defined in this paper as "how native English speakers perceive or interpret errors made by EFL learners." Japanese learners of English sometimes make such an error as "How do you feel her?" when their intended meaning is "How do you feel about her?" In this case we assume that it would be more helpful for the learners if they are presented with its error perception and interpretation ("How do you feel her?" means " 'How do you touch her?' in a sexual meaning.") than if they are only given its correct expression ("How do you feel about her?"). We predict if learners know how native English speakers perceive and interpret the errors which learners make, there is a strong likelihood that they will not make the same errors thereafter. Katagiri (1997b) investigated the efficacy of presenting all the learners in intact classes with the information on error perception and interpretation, and no statistically significant positive effect was found. The present research investigated the efficacy of presenting only the learners who made errors of which possible perceptions and interpretations by native speakers the instructors gave. No statistically significant positive effect of presenting the learners with information on perceptions and interpretations of their actual errors was found.
