Martha E. Crosby

How do we read algorithms? A case study

The authors report on an experiment that explored the way subjects view an algorithm, written in Pascal, and the graphical representation of its behavior. They designed the experiment to examine the influence of programming experience on how subjects view a short but complex algorithm, and they analyzed the scan patterns of eye movement for subjects at two experience levels. They describe the experimental procedure in detail, and discuss the results. An important finding was that all subjects needed numerous fixations in most areas of the algorithm and spent more time viewing relevant slide areas than did subjects in studies using simple text. This supports the hypothesis that viewing strategies for algorithms differ from those for prose.<<ETX>>

Assessing Cognitive Load with Physiological Sensors

Assessing the cognitive load of a subject performing a computer task using task performance data is normally available at the end of the task. For assessing cognitive load, physiological data has the advantage of being available in real-time and the potential of assessing the affective components of cognitive load. Described are two new methods of assessing cognitive load from eye tracking and the pressures a subject applies to a computer mouse when subjects perform a math task that involves moving targets. Physiological measures that significantly discriminated task difficulty included eye movement, skin conductivity and one of the pressure signals from the computer mouse. Also, in some cases, these physiological measures can be more sensitive than task performance measures of cognitive load (i.e., incorrect actions) to detect interaction effects with task difficulty. The suite of physiological sensors is shown to be a viable alternative or supplement to task performance measures.

The effect of graphic style on data model interpretation

Abstract. Graphic data models are commonly used as a tool for presentation of information structures in the design, implementation, use and maintenance of the databases that support information systems. The methods proposed for database design assume that the use of graphic data models will enhance understanding of system specifications by both the end‐users and the implementers of the system. For this assumption to hold, the information presented in the graphic data model must be readily comprehensible so that the design, represented by the model, can be confirmed and implemented correctly. The lack of standard representations for graphic models has led to a variety of graphic styles. To date, there has been little focus on studying the effect graphic style has on model comprehension. We have studied the effect of three graphic styles proposed for data models on model legibility and interpretation. Our study shows a significant variation in model interpretation that can be attributed to the graphic syntax used. Graphic style appears to influence which model elements are included in the interpretation, as well as the way data models are read.

Eye movements in code reading: relaxing the linear order

Code reading is an important skill in programming. Inspired by the linearity that people exhibit while natural language text reading, we designed local and global gaze-based measures to characterize linearity (left-to-right and top-to-bottom) in reading source code. Unlike natural language text, source code is executable and requires a specific reading approach. To validate these measures, we compared the eye movements of novice and expert programmers who were asked to read and comprehend short snippets of natural language text and Java programs. Our results show that novices read source code less linearly than natural language text. Moreover, experts read code less linearly than novices. These findings indicate that there are specific differences between reading natural language and source code, and suggest that non-linear reading skills increase with expertise. We discuss the implications for practitioners and educators.

Physiological data feedback for application in distance education

This paper describes initial experiments collecting physiological data from subjects performing computer tasks. A prototype realtime Emotion Mouse collected skin temperature, galvanic skin response (GSR), and heartbeat data. Possible applications to distance education, and a second-generation system are discussed.

Reducing cognitive load

This paper explores issues related to cognitive load in the contexts of learning, information filtering, user modeling, categorization and personal information organizing behavior. We incorporate expertise from the fields of information science, educational psychology and computer science to report research that can ultimately influence the design of personalized adaptive systems.

A model for integrating an adaptive information filter utilizing biosensor data to assess cognitive load

Information filtering is an effective tool for improving performance but requires real-time information about the users changing cognitive states to determine the optimal amount of filtering for each individual at any given time. Current research at the Adaptive Multimodal Interactive Laboratory assesses the users cognitive ability and cognitive load from physiological measures including: eye tracking, heart rate, skin temperature, electrodermal activity, and the pressures applied to a computer mouse during task performance. A model of adaptive information filtering is proposed that would improve learning and task performance by optimizing the human-computer interface based on real-time information of the users cognitive state obtained from these passive physiological measures.

Japanese and American Students Meet on the Web: Collaborative Language Learning through Everyday Dialogue with Peers

This case study discusses the design, implementation, and results of a series of collaborative Japanese language-learning sessions conducted synchronously and asynchronously via the Internet among globally distributed, cross-institutional teams of students. MUD/MOO-like team-room scenarios required that students engage in more extensive use of language skills than would be required in usual classroom work. This virtual classroom design provided a naturalistic environment for computer assisted language learning (CALL) while motivating students to adapt more readily to interaction with peers in a varied, complex, and uncertain environment. At the same time, the tasks to be accomplished provided students with the opportunity to engage in virtual teamwork with peers in a challenging and enjoyable problem-solving and decision-making context. Within this framework, AI techniques that enhance team memory, sustain workflow processes, and employ user/learner models will play an imp ortant role in the near future.

User identification based on the analysis of the forces applied by a user to a computer mouse

This paper describes the framework for a branch of augmented cognition research performed at the Adaptive Multimodal Laboratory at the University of Hawaii and a spec application involving the identification of a computer user based on the forces applied to a computer mouse (i.e., click signature) during a task. Data was collected from six people during a pilot study. Two methods used to identify users were a back propagation neural network and discriminant analysis. Results indicate that the discriminant analysis was slightly better at identifying users than the neural network, but its primary advantage was that it required less data preparation. Continuous identification of the user is possible with either method. Successful, identification of the user is a useful first step to proceed to the next stage of the research framework, which is to identify the users cognitive state for implementation in an augmented cognition system.

What Eye Fixation Patterns Tell Us About Subitizing

Differences between the enumeration of very small (1–3) versus larger (4–6) numerosities were examined by investigating where people fixate when they are enumerating different numbers of items. Overall, fixations were more likely to be located in regions of the array that contained target items when the array contained 4 or more targets than when it contained 3 or fewer, a result that is consistent with previous research indicating that the enumeration of very small sets is less dependent on attentional processing than is the enumeration of larger sets. However, both the pattern of fixations across different distractor conditions and an analysis of the temporal course of fixations in the absence of distractors were inconsistent with a dichotomous distinction between pre-attentive and attentional forms of enumeration. Rather, our results suggest that, irrespective of numerosity, enumeration entails some attentional processing of target items, but attentional processing plays a markedly greater role in the enumeration of 4 or more items than in the enumeration of 3 or fewer.