This article provides an overview of the EPIC architecture being developed by Kieras and Meyer for modeling human cognition and performance (Kieras, Wood, & Meyer, 1997; Meyer & Kieras, 1997a, 1997b). EPIC is similar in spirit to the Model Human Processor (MHP; Card, Moran, & Newell, 1983), but EPIC incorporates many recent theoretical and empiri- cal results about human performance in the form of a software framework for computer simulation modeling. Using EPIC, a model can be con- structed that represents the general procedures required to perform a complex multimodal task a a set of production rules. When the model is supplied with the external stimuli for a specffic task, it will then execute the procedures in whatever way the task requires, thus simulating a human 's performing the task and generating the predicted actions in simulated real time. EPIC is an architecture for constructing models of 394 performance. It is notet a learning system and so has no mechanisms for learning how to perform a task. Rather, the purpose of EPIC is to repre- sent in detail the perceptual, motor, and cognitive constraints on the human ability to perform tasks

Engineering models of human performance permit some aspects of usability of interface designs to be predicted from an analysis of the task, and thus can replace to some extent expensive user testing data. The best developed such tools are GOMS models, which have been shown to be accurate and effective in predicting usability of the procedural aspects of interface designs. This paper describes a computer-based tool, GLEAN, that generates quantitative predictions from a supplied GOMS model and a set of benchmark tasks. GLEAN is demonstrated to reproduce the results of a case study of GOMS model application with considerable time savings over both manual modeling as well as empirical testing. KEYWORDS User-Interface Software and Technology: Usability, usability evaluation, user models, GOMS models INTRODUCTION Engineering Models for Usable Interface Design The standard accepted technique for developing a usable system, empirical user testing, is based on iterative testing and design r...

this article, NGOMSL, in which learning time and execution time are predicted based on a program-like representation of the procedures that the user must learn and execute to perform tasks with the system. NGOMSL is an acronym for Natural GOMS Language, which is a structured natural language used to represent the user's methods and selection rules. NGOMSL models thus have an explicit representation of the user's methods, which are assumed to be strictly sequential and hierarchical in form. The execution time for a task is predicted by simulating the execution of the methods required to perform the task. Each NGOMSL statement is assumed to require a small fixed time to execute, and any operators in the statement, such as a keystroke, will then take additional time depending on the operator. The time to learn how to operate the interface can be predicted from the length of the methods, and the amount of transfer of training from the number of methods or method steps previously learned. Thus estimating times for learning and execution both require counting the number of NGOMSL statements involved; details on this process will be provided in this article.

To understand how people search for a known target item in an unordered pull-down menu, this research presents cognitive models that vary serial versus parallel processing of menu items, random versus systematic search, and different numbers of menu items fitting into the fovea simultaneously. Varying these conditions, models were constructed and run using the EPIC cognitive architecture. The selection times predicted by the models are compared with selection times of human subjects performing the same menu task. Comparing the predicted and observed times, the models reveal that 1) people process more than one menu item at a time, and 2) people search menus using both random and systematic search strategies.

We report an investigation into the processes involved in a common graph reading task using two types of Cartesian graph. We describe an experiment and eye movement study, the results of which show that optimal scan paths assumed in the task analysis approximate the detailed sequences of saccades made by individuals. The research demonstrates the computational inequivalence of two sets of informationally equivalent graphs and illustrates how the computational advantages of a representation outweigh factors such as user unfamiliarity. We describe two models using the ACT rational perceptual motor (ACT-R/PM) cognitive architecture, that replicate the pattern of observed response latencies and the complex scan paths revealed by the eye movement study. Finally, we outline three guidelines for designers of visual displays: Designers should (a) consider how different quantities are encoded within any chosen representational format (b) consider the full range of alternative varieties of a given task, and (c) balance the cost of familiarization with the computational advantages of less familiar representations. Actual or potential applications of this research include informing the design and selection of appropriate visual displays and illustrating the practice and utility of task analysis, eye tracking, and cognitive modeling for understanding interactive tasks with external representations.

Abstract not found

Unistrokes are a viable form of text input in pen-based user interfaces. However, they are a very heterogeneous group of gestures the only common feature being that all are drawn with a single stroke. Several unistroke alphabets have been proposed including the original Unistrokes, Graffiti, Allegro, T-Cube and MDITIM. Comparing these methods usually requires a lengthy study with many writers and even then the results are biased by the earlier handwriting experience that the writers have. Therefore, a simple descriptive model can make these comparisons easier. In this paper we propose a model for predicting the writing time for an expert user on any given unistroke alphabet thus enabling sounder argumentation on the properties of different writing methods. Keywords Modeling of motor performance, handwriting, pen input INTRODUCTION Unistrokes were introduced as a text input method for penbased user interfaces by Goldberg and Richardson in their 1993 paper [8]. Unistrokes are an alte...

Abstract: Domain-Specific Visual Modeling Languages should provide notations and abstractions that suitably support problem solving in well-defined application domains. From their user’s perspective, the language’s modeling primitives must be intuitive and expressive enough in capturing all intended aspects of domain conceptualizations. Over the years formal and explicit representations of domain conceptualizations have been developed as domain ontologies. In this paper, we show how the design of these languages can benefit from conceptual tools developed by the ontology engineering community. 1.

Abstract: To justify use of a simulation modeling framework in designing a human-machine system, engineers need to have some reasonable expectation that doing so will prove beneficial. For developers of modeling frameworks this presents two main challenges. First, the framework must be able to predict design-relevant aspects of human performance – i.e. it must be able to make predictions that designers care about. Second, the time and expertise required to use the framework must be kept within practical limits. This paper will describe efforts to address these problems using the APEX modeling framework. When designing complex devices such as integrated circuits and automobile engines, engineers routinely use computer simulation to predict how well the device would function if actually built. By helping to detect problems at an early stage in the design process, simulation postpones or eliminates the need for a physical prototype. Engineering costs decrease in numerous ways resulting in improved reliability, greater innovation, faster development time, and lower overall cost of development. While a routine part of the design process for some devices, simulation is hardly ever used to help design human-machine systems. There are two main reasons for this, each stemming from difficulty modeling the human components of these systems. First, available frameworks for modeling human performance often prove unsatisfactory because they are incomplete in some crucial way, or cannot adequately predict design-relevant aspects of human performance. Second, using such a framework requires spending a great deal of time and effort preparing application-specific elements of the simulation including models of newly designed devices and formal descriptions of “how-to” knowledge for operating in the domain of interest. The limitations of most human models often make it unlikely that this investment of time and effort will pay off. Despite these limitations, human operator models have occasionally proven effective in

This paper builds on an existing model of human-computer interaction that focuses on the information structures employed as distributed information resources by users in deciding how to act. Such resources are not limited to internal mental constructs, but may be distributed between the single user and their environments, as representations in artefacts or the heads of other human actors. The concept of interaction strategy is introduced and used as a theoretical construct that generalises over a number of modelling perspectives in HCI. They are used to form a link between users' high-level objectives and resources and the way they're implemented. The usefulness in design of the general idea of distributed information resources and the concepts of strategies and objectives is illustrated by an example. KEYWORDS Interaction design, interaction strategy, distributed cognition 1. INTRODUCTION This paper uses a framework in which alternative theoretical perspectives on HCI modelling pro...