Journal of Educational Computing Research

Computer-based learning environments have been defined as settings where students use software to facilitate active, exploratory learning. We distinguish between illocutionary and non-illocutionary learning environments: the computer interprets the intent of students in the former but not in the latter. We apply these terms to language learning environments, and report descriptive research on ESL students' use of a non-illocutionary learning environment. Results indicated students “explored” the learning environment in a routine way, but failed to explore creatively the program's morphosyntactic possibilities; routine exploration was positively related to attitudes for part of the group, but negatively related to attitudes for students who may have found the software environment too easy. Implications for illocutionary and non- illocutionary learning environments are discussed.

The research project “Feedback Processes in Computer Managed Spelling Instruction” was aimed at developing and testing an instructional software program for teaching the orthography of Dutch verbs. The main focus of this article is on how to design an optimal learning environment. The design of human-computer interaction is highlighted from a number of respects: the sequencing of the learning content, the presentation of information on the screen, and the format of student input and feedback. The program consists of seven instructional modules. In each module, a specific grammatical operation, such as grammatical tense, person, number or voice, is highlighted. The effectiveness of the courseware was tested in a field experiment. Two versions of the program, a Standard Version and a Flexible Version, were compared, taking account of individual differences in spelling ability and motivation variables. Results with both versions are reported.

High school students enrolled in precalculus/trigonometry classes were randomly assigned to receive instruction either by a master teacher in a regular classroom or via interactive videodisc instructional modules. The interactive modules were designed to help students understand mathematical concepts by means of real-world video exemplars of the concepts, and by modeling problem-solving strategies and metacognitive processes. Evaluation results showed that students who learned solely by means of the interactive videodisc materials achieved as well or better than students who received traditional instruction organized around the same objectives. The relation of student goal orientations and self-perceptions to learning outcomes was also investigated. The results of multiple regression analyses for each of eight instructional topics revealed a pattern supportive of Dweck's theory regarding the role of goal orientations in academic motivation.

The computer inclination (CI) of 881 eleventh-grade students and of their 181 teachers in relation to their Science-Technology-Society (STS) views was assessed in randomly selected high schools in British Columbia using a two-part questionnaire: a short, modified version of the Survey of Attitudes Toward Learning About and Working With Computers for the CI, and a condensed form of the Views on Science-Technology-Society (VOSTS) Inventory for the STS views. Significant differences between the computer inclinations of non-STS and STS students and between the STS views of non-STS and STS teachers and students in the “attitude” category were found, but there were no differences between other subgroups on these dimensions. Computer Inclination Indices (CII) and Science/Technology Indices (ST_x) have been defined and used for the elucidation of possible correlations between the CIIs and the ST_xs. It is argued that the CIs and STS views of both students and teachers should be assessed and taken into consideration by the educational policy makers and developers of future science curricula and teacher training programs, if the goals of the currently sought reforms in science education are to be attained.

The two purposes of this study were a) to explore how young children structure and solve geometric analogy problems in a computer-based test, and b) to compare this performance to the same item stems in a paper-and-pencil form. Twenty-four kindergarten and twenty-four second-grade children were required to construct solutions to analogy problems that were presented on a microcomputer. Half of the sample was academically gifted. These kindergartners were also tested with a paper-and-pencil version of the test to examine for possible differences between the two testing methods. Second graders were only tested with the computer version. Statistical analysis for the computer-based test focused on how these young children constructed responses to geometric analogy problems, and the error patterns committed by each age group. Analyses showed a strong linear decrease in performance as the item information load increased, but this effect was more pronounced for the kindergartners than for the second graders. Analyses also provided some evidence that changing the items from a paper-and-pencil response selection format to a constructed response format alters which component of the item carries the greatest weight in determining overall item performance. This study also investigated the crossover effects of computer-based testing on paper-and-pencil testing and vice versa in the kindergarten subjects. It was found that there was a positive effect on the post-test computer-based performance when pretested with the paper-form, but a negative effect on paper-form post-test performance when pretested with the computer form.

A study was undertaken to determine the feasibility of using a computer program to mark examinee write-in responses. It compared the performance of a computer against that of physician markers. A new examination designed to assess clinical decision-making was administered to eighteen residents. The write-in responses on twenty-three questions were marked both by physicians and a microcomputer. The study found the computer to be more consistent in marking, equally as cost effective, and as valid as physician markers. Since the computer provided question statistics (i.e., means, frequencies of examinees' responses) it was evaluated to be better and more helpful than physician markers by providing data useful for improving examination questions and learning experiences.

This article discusses the concept of a microworld and the role it might play in education. Three questions are asked of previous descriptions of microworlds: what does the microworld look and feel like, what are its design features, and what sort of learning happens? It is argued that most microworld designers begin from a desire to teach particular ideas, and design for an “exploratory” mode of learning in which the student explores the microworld to find hidden nuggets of knowledge. An alternative design strategy is described in an account of a microworld intended to help improve understanding of proportion. The design of this microworld, which arose from a commitment to helping with a specific learning difficulty, was based on research about the nature of that learning difficulty, and was designed for an “expressive” mode of learning in which the student is provided with tools to express his/her own ideas.

A common assumption made by teachers of programming languages, especially those teaching adults, is that their primary (core) teaching material consists of text, by which we mean rules, instructions, explanations, and discussions. Worked examples are seen as secondary, serving merely to illustrate the core material. We present evidence to demonstrate that young adult students do not share this assumption. Given instruction in the form of demonstrably clear text and worked examples, they do not treat the text as the primary source of information, but rather seem to focus on the worked examples. Worked examples may be seen as being the primary source of learning material even when they are not fully understood, but bear some superficial similarity to the task in hand.