G. Michael Barnes

The Implications of Processing Event Sequences for Theories of Analogical Reasoning

Sternberg (1977a, 1977b) has proposed a componential theory of information processing on analogies. The current study attempts convergent validation of the basic findings in verbal analogies by a method that is based on different underlying assumptions. Although the data were generally consistent with Sternberg’s theory, the data indicated thatapplication is better described by two separate events. Furthermore, the extent of individual differences in strategy models was so substantial that a higher level processing operation, such as control strategies, should be postulated, rather than to support a single-strategy model as characterizing Analogy solving.

Problem restructuring processes for ill-structured verbal analogies

The processes involved in analogy solving have been an important investigative area in cognitive psychology. Although problem restructuring has been a central construct in problem solving theory, no restructuring processes have been proposed for analogical reasoning. Yet, the stimulus terms for analogies, as they appear on ability tests, are often ill-structured. That is, they are ordered in a way that does not permit direct problem comprehension. In the current study, both perceptual and semantic problem restructuring processes were hypothesized for analogy solving. The independence, stage of execution, and susceptibility to strategic control of the two processes were examined. The results from two experiments indicated that (1) ill-structured analogies are restructured during problem solving, (2) perceptual and semantic restructuring processes are independent and executed at different stages of analogy solving, and (3) both processes exhibited automaticity since repetition of analogy solution attenuated but did not eliminate either restructuring process. A model of analogical reasoning that incorporated both restructuring processes and their execution sequences was proposed. The nature and automaticity of perceptual and semantic analogy restructuring processes were disscussed.

A computer science courseware factory

A model for computer science courseware development within universities is proposed. This model asserts that masters level graduate and undergraduate students are a valuable software development resource that can be utilized to design, prototype, field test, and refine quality courseware. To support this contention four prototype courseware packages are used to briefly describe the Visible Algorithms project at Northridge. These packages provide tutorials on, and graphic animations of: linked list, binary tree, B-tree, and memory management algorithms.

Experiments with balanced-sample binary trees

In this paper we propose using experiments with Balanced-Sample Binary Trees (BSBTrees) as assignments and lecture material in intermediate data structures courses (CS2/3). BSBTrees are composite data structures that have a temporarily constructed form that precedes their normal construction. We present them in the context of binary search trees. To do this we first investigate the retrieval properties of randomly generated binary search trees and show how temporary construction can improve both worst case and average case behavior. We provide a brief analysis of BSBTree performance and description of the classes that can be used for BSBTree implementation. Last we discuss the use of BSBTrees in CS2 and CS3 courses and a survey of student opinions about BSBTrees.

Visual simulations of data structures during lecture

A microcomputer based software package for the graphic simulation of fundamental binary search tree algorithms is presented. This application was developed to augment instructional aids available to the instructor during classroom lecture. The system is mouse controlled and was designed to be used on an IBM PC compatible microcomputer connected to either a large screen video monitor or a projection display. Several screen dumps are presented to illustrate the multi-level graphical representation of trees, pseudocode algorithms, and window environment incorporated in the package. In addition, the literature on CAI oriented simulations for Computer Science topics, with a focus on data structures, is reviewed briefly. Computer Aided Instruction (CAI) is an educational application area with tremendous potential. The recent advent of reasonably priced personal computers with medium resolution color graphic capabilities will hopefully increase the generation of CAI materials in all fields. Ironically, there has been little CAI material developed to aid in the instruction of Computer Science. Furthermore, earlier efforts at courseware development for computer science was centered on programming languages not more advanced topics. In recent years, however, several researchers have been developing personal computer based courseware simulations of more advanced topics: program execution [2,6,21,16], computer architecture [15,17], and data structures algorithms. This work is part of a growing body of research being performed in the emerging area of Visual Programming [11,10]. In this paper we will first briefly review the area of graphic simulations of fundamental data structures and their algorithms. Second, we will describe the CABTO (Computer Animation of Binary Tree Operations) package. We developed CABTO to be used as a lecture aid in a lower division data structures course.

Experiments with file accessing techniques

This paper describes a series of experiments designed to be carried out by students in an introductory file processing class. The experiments were class assignments that involved student teams implementing various file accessing techniques. For each implementation, teams had to run a set of benchmark tasks and obtain software metrics. In addition each team was required to write a report analyzing the results of their experimentation. Our four objectives were to provide students with experiences in: file processing, team programming dynamics, generating software metrics, and the experimental evaluation of different solutions to a set of problems.

Creativity: A complementary relationship between information demand and success at problem solving.

Abstract Creativity, intelligence, and problem-solving tasks that varied the explicitness of problem definition and the format for an acceptable solution were administered to 100 subjects. Creativity and intelligence tasks were described as different problem-solving tasks. Two hypotheses that distinguished between concepts of creative and intelligent problem solving were presented and supported. First, the syntactical hypothesis stated that the format of an acceptable solution would define the factor structure of the problem-solving tasks more parsimoniously than the explicitness of the problem definition. The factor structure was interpreted by the type of acceptable solution format. Second, the strategic hypothesis stated that information demanded to define a problem and success at problem solving would interact. Information demanded and success at problem solving did interact producing a complementary relationship with dimensions of creativity but not intelligence. Thus, syntactical and strategic distinctions between creativity and intelligence were obtained.

Effects of Scroll Bar Orientation and Item Justification When using List Boxes

List boxes are a common user interface component in graphical user interfaces. In practice, most list boxes use right-oriented scroll bars to control left-justified text items. A two-way interaction hypothesis favoring the use of a scroll bar orientation consistent with list box item justification was examined for speed of use and user preference in two experiments. A two-way interaction was obtained in both experiments. Item selection was faster with a scroll bar orientation consistent with list item justification. Subjects preferred right-oriented scroll bars with right-oriented text items. There was no difference in preference between right and left-oriented scroll bars for left-justified text items. These results support a design principle of locality for user interface controls and controlled objects.

Digitized speech's serial position effect

Speech is becoming a more important user interface with the advent of multimedia capable video game systems, personal computers and workstations. Speech can be synthetically generated or natural speech can be digitized for storage and replay. Several studies have shown that synthetic speech is not understood as well as natural speech [2]. Synthetic speech lacks prosodic (timing) and acoustical cues to phonetic segments in comparison to natural speech. Luce et. al. [2] attributed the difficulty with synthetic speech to increased processing time requirements to encode the degraded sound. Waterworth and Thomas [3] replicated Luce et. al. serial position effect experiment and reported additional support for the encoding hypothesis by having subjects repeat words. When they adjusted for correct word encoding, the synthetic speech was similar to natural speech.