No abstract

Short-term memory for Chinese radicals and characters, varying in orthographic complexity, frequency, and-for radicals-intercharacter frequency (the number of compound characters that contain the radical), was studied using an immediate free-recall task. When radicals or characters are relatively frequent, so that their pronunciations are well known by literate Chinese, they seem to be maintained in verbal form in short-term memory. For these stimuli, intercharacter frequency and complexity have relatively small influences on memory span. Stimuli low in frequency, with pronunciations that are not apt to be known, seem to be maintained in visual form in short-term memory. Memory span is much smaller for these stimuli and is influenced by both intercharacter frequency and complexity. Furthermore, short-term memory for relatively highfrequency characters is interfered with more by a verbal than by a visual intervening task, whereas the opposite is true for low-frequency characters.

The development of information and communication technology changes how, what, who, when, where and why we learn. Unfortunately, little is known of the exact impact that these changes will bring to education. However, we are certain that many new learning and teaching styles which are called learning models in the paper will emerge to cope with the changes in the near future. The present paper describes four spaces of learning models, namely, the future-classroom, the community-based, the structuralknowledge, and the complex-problem learning models, which are specifically designed to integrate the Internet into education.1 With the four spaces of learning models, the present paper may serve two functions. First, it offers a way to integrate an array of different communication technologies (e.g. handheld computer, wireless communication and the Internet) and learning theories into an integrated schema. Secondly, the paper offers a direction concerning how and what to look for in education with the Internet integrated in. #

No abstract

Subjective ratings of action-sequence strength and scene structure were collected for the 30 scripts documented by Galambos (1983). Subjects largely agreed on the location of the major scene boundaries. Ratings of action sequences indicated that some scripts possess stronger sequential properties than others, and if actions were from the same scene rather than from different scenes, the probability that subjects disagreed about their order was higher. We discuss the implications and importance of these results concerning modeling of script representation in memory.A script has been defined as a "predetermined, stereotyped sequence of actions that define a well-known situation" (Schank & Abelson, 1977, p. 41). During the past two decades, much research has been devoted to the functional aspect of scripts, using a variety of tasks such as reading (e.g.,