Learning to read Chinese with computer-based speech feedback

Two experiments were conducted to determine the optimal computer-based reading instruction method for readers beginning to learn Chinese. In Experiment 1, 24 Hong Kong Primary One students practised reading Chinese texts in three different conditions. The read-along condition visually displayed the whole text while having it vocalized at a normal reading rate by the computer. The read-along-with-cue condition resembled the read-along condition except that visual character-by-character highlighting was also available in which the timing matched that of the vocalization. In the speech-at-will condition, the text was presented with character-by-character highlighting but the vocalization was available only when it was requested by the reader. The reader was tested on word naming both before and after the training. The learning effect of the read-along-with-cue condition was greater than that of the speech-at-will condition. Experiment 2 was a replication of Experiment 1, using new texts of various contexts. The read-along-with-cue condition was superior to both the speech-at-will and the read-along conditions. However, most learners preferred using the speech-at-will format. The current results indicate that the read-along with cue technique is the best practice condition For word identification and that there is a clear dissociation between the most preferred and the most effective training method. Copyright © 1997 The Chinese University of Hong Kong.