Contents of Volume 30 2002

Volume Contents

Contents of Volume 30 2002     

The application of machine learning to student modelling

This article considers the prospects for applying machine learning techniques to the problem of maintaining the dynamic student models needed for intelligent tutoring. We imagine a learner endeavouring to understand a climate classification scheme through exploring a data-base guided by a tutorial system which does not itself know the classification rules. We consider the possible use of the focussing algorithm for concept learning as a basis for such a tutorial system. Several difficulties with this approach to guided discovery learning are discussed.     

Work analysis by random sampling

Random sampling of work activities using an electronic random alarm mechanism provided a simple and effective way to determine how time was divided between various activities. At each random alarm the subject simply recorded the time and the activity. Analysis of the data led to reassignment of staff functions and also resulted in additional support for certain critical activities.     

Initial mental graphemic representation acquisition and later literacy achievement in children with language impairment: a longitudinal study

The purpose of this study was to examine the relation between the ability to quickly acquire initial mental graphemic representations (MGRs) in kindergarten and fourth grade literacy skills in children with typical language (TL) and children with language impairment (LI). The study is a longitudinal extension of a study conducted by Wolter and Apel in which kindergarten children with LI and TL were administered early literacy measures as well as a novel written pseudoword task of MGR learning (spelling and identification of target pseudowords). In the current study (4 years later), the authors administered reading and spelling measures to 37 of the original 45 children (18 children with LI, 19 children with TL). The children with LI performed significantly lower than their peers with TL on all fourth grade literacy measures. For both groups, kindergarten initial MGR acquisition ability significantly related to fourth grade real-word reading and spelling. For the children with LI, kindergarten initial MGR acquisition ability also related to fourth grade pseudoword decoding and reading comprehension. Collectively, the findings suggest that initial MGR learning in kindergarten is an essential skill that may uniquely relate to later literacy abilities.     

Beyond intelligent tutoring systems: Situations,interactions, processes and affordances

The idea of an intelligentsystem to support learning has been epitomisedby that of an intelligent tutoring system(ITS). However, ITSs are, in fact, just aparticular kind of intelligent system tosupport learning whose components reflect thevalues of the particular view that ITSsemphasise in regard to the nature of knowledge,learning and teaching, which have led to anarchitecture that focuses on representing theknowledge to be learned (domain model),inferring the learner's knowledge (learnermodel), and planning instructional steps to thelearner (teaching model). On the other hand,other views of learning may lead to differentneeds in terms of knowledge representation,reasoning, and decision making capabilities inthe intelligent systems that support them.Constructivist views, for example, emphasisedifferent values and may require an entirelydifferent architecture of intelligent system tosupport its philosophy of learning. This paperpresents an architecture of an intelligentsystem to support learning that is able toaddress the issues that arise fromconstructivist theories of learning in a waythat, rather than opposing to the standard ITSarchitecture, characterises a broader view inall its components which can be appropriatelyattuned to address the issues of particularphilosophies.     

Preschool outcomes of children of depressed mothers: role of maternal behavior,contextual risk,and children's brain activity

Children of depressed mothers are at risk for behavioral and emotional problems. Infants of depressed mothers exhibit behavioral disturbances and atypical frontal brain activity. The mechanisms by which children develop such vulnerabilities are not clear. Three-year-old children of mothers with (N = 65) and without (N = 59) a history of depression were assessed in terms of behavior problems and brain electrical activity. Children of mothers with chronic depression exhibited lower frontal and parietal brain activation compared with children of mothers without depression and those whose depression remitted. Depressed mothers reported higher contextual risk (e.g., marital discord and stress) and their children had more behavior problems. Children's frontal brain activation and contextual risk level mediated the relation between maternal depression and child behavior problems.     

Intelligent educational systems: identifying and decoupling the conversational levels

Intelligent educational systems (IESs) have traditionally been thought of in terms of their functional subsystems. The three obligatory subsystems are those containing domain knowledge, a model of the learner's current state, and teaching knowledge. We argue that it is useful to reconceptualise IESs: we should think of them from the point of view of the multiple levels of conversational interaction that they support with a learner. We suggest that in many cases it is useful to identify a task level at which the learner is working, and a higher discussion level at which the learner and system converse about, and reflect on the primary task. If the two levels can be to some extent decoupled, meaning that each level has less than complete knowledge of the other level, IES development might be simplified. In some cases an IES might be built by the addition of a discussion level component to an existing computer tool, simulation, game or microworld. A critical requirement for a decoupled IES is a representation for communication between levels; study of human tutoring is needed to guide identification of suitable representations.