Abstract
In this study, students with developmental disabilities were trained to answer the cause of emotional states expressed by a character in a printed passage. In the passage, daily events happen and then, the character expresses either "crying" or "laughing" emotional states (e.g., John went for a walk the park. John met a big dog and was bitten by the dog. John's leg was bleeding. John is crying) . After the passage was read by the students,the trainer asked the cause of the emotional states (i.e. "Why is John crying?") . In the initial intervention, story-emotion words (i.e. "hurt", "sad", "tickle", "funny") matching-to-sample procedure was used. However under these training conditions, it was not possible to show generalization in untrained tasks so that both students needed additional training to choose appropriate emotion words. After this training, the students were able to answer respective Test tasks with appropriate emotional expressions in Probe 1. Next, the events passage which triggered the emotional states were underlined and the matching-to-sample procedure was used again. After that, the students were able to answer not only with appropriate emotional expressions but also with answers which involved the cause of the emotional states (e.g. "Because the dog bit and it hurt"). The answering skills were generalized to other untrained tasks in both students. Results are discussed from a behavior analytic view about the effectiveness of conditional discrimination training for reading comprehension and the importance of making the contextual stimulus stand out.
