Search

This paper describes a methodology for exploring trust using psychological (subjective) and physiological (objective) correlates to trust. The aim was to explore trust using natural dialogs of real-world scenarios that embed fifteen subjective measures. The goal was to apply the method in modeling human-robot-human interaction, involving three types of androids and to predict trust. Two forms of dialogs were employed: a guided script and a predetermined dialog representing three social scenarios. Objective features included facial expressions, voice and heart rate. Subjective trust measures comprised ability, benevolence and integrity. A repeated measures experimental design was employed. Forty-two subjects participated in the study. The data was analyzed using exploratory factor analysis and correlation. Multiple neuro-fuzzy models were trained using the data set and combined as an ensemble using evolutionary algorithms. The final ensemble estimated trust with 67% accuracy. The implications of the findings and limitations of the method are discussed.

Information access cost has been reported to be one determinant of one’s information retrieval strategy. However, the “expected gain” of an information retrieval strategy has not received much attention. The present study aimed to examine the effects of confidence judgment and overconfidence bias, besides the information access cost, on the choice of information retrieval strategy and task performance. The results showed that both information access cost and confidence judgment of memory accuracy affected information access attempts, and overconfidence bias harmed information retrieval performance.

Global workplaces and multinational organizations in the oil and gas industry have created an environment in which human factors/ergonomics professionals collaborate to solve office ergonomics and process control design problems for clients. The demand for ergonomics expertise is growing, but the supply of certified ergonomists is limited. The situation is acute in Southeast Asia (SEA), given the lack of ergonomics awareness, training, and certification. We present three challenges that required ergonomics interventions and collaboration among ergonomists. Two of the projects involved multinational companies operating in SEA and one, a national company with global operations.

Interruptions during surgical counts are critical gateways for adverse events to happen. It is important to understand the cognitive strategies nurses employ during interruptions of counts and to r...

This study explored whether explicitly presenting rate-of-change cues, an information that is otherwise only implicitly derived in today’s process control operations, would benefit operator performance. Operators rely on rate-of-change (ROC) information to help anticipate and arrest problems and take early action. Currently ROC is derived implicitly via Trends displays, which plot historical data over time. A simulated experiment conducted using university undergraduates presented explicit rate-of-change (ROC) information, either in numerical or linear-shape format, into a modified Trends displays. Findings revealed performance improvement patterns and particularly strong performance benefits when ROC was presented in linear-shape format. Results provide first step towards integrating ROC representation into industry displays.

Introduction: In a cognitive system, human operators work with automation to achieve their goal. However, there are times when the human operator cannot interpret what the automation is doing. This can create dangerous communication gap particularly if the automation provides the wrong advice. This paper proposes the Motivation-Expectation Space (MES) as a representative structure for bridging the gap. Review: The MES has its theoretical foundation from research in Cognitive System Engineering. It represents the goals, means, causes, and effects in an orthogonal structure. These four components were repeatedly observed in research related to decision strategies. The MES was assessed against the information requirements for improving automation trust and dependency and found to suffice in supporting most of the requirements. Contribution: The structured representation of these elements in the MES would likely facilitate the interpretation of the decision strategies, making it easier and faster for the human to understand what the automation is doing. Parallels were also drawn between MES and Situation Awareness to suggest the potential value MES can bring to research in Cognitive System Engineering.

Proactive monitoring involves, in part, the effective use of time-based information such as Trends. In the oil and chemical industries, the benefits of a proactive approach include less down-time, fewer alarm-flooding stresses, and maximized product output. This report aims to describe the effective use of trend displays by examining how expert Distributed Control System (DCS) operators monitored and visualized trends proactively. We found two preferences for trend display layouts, suggesting a lack of design guidelines and understanding towards optimum trend usage performance in the industry. The decision process in proactive monitoring is discussed, along with the attributes that define an expert trends user.

Emotion is elicited by products upon closer inspection before purchase. The pre-purchase situation is very important because it is at this stage that consumers make a decision to buy a product. This study identified the relationship between mobile phone attributes and pre-purchase emotion. The moderating effect of selected purchase criteria to the relationship between attributes and emotion was also be investigated. Forty two participants voluntarily took part in this field survey. Results of the study suggested that phone attributes related to dimension and form can predict pre-purchase emotion. The size of the navigation button was found to be a significant predictor of all pre-purchase emotion except excitement. Larger navigation buttons elicited greater intensities of positive emotion. Other phone attributes were also found to be significant predictors of some emotions. Design preference moderated the relationship between the size of the navigation button and feelings of encouragement to buy.

This paper describes a method for transforming knowledge elicited from real life applications into Rasmussen's Abstraction-Decomposition Space (ADS). The ADS is used to represent knowledge in a Work Domain Analysis (WDA), the first of five stages in a Cognitive Work Analysis. The WDA is highly conceptual and is mostly used by trained practitioners. There are many concepts and rules in the construction of the ADS. These make it difficult for practitioners to transform and map their data onto the two dimensional (2D) ADS. In Applied Work Domain Analysis (AWDA), we introduce a new dimension, cause-effect, thereby creating a 2D orthogonal space called the Motivation-Expectation Space (MES). The rules for constructing the MES are easier to apply than the ADS. After constructing the MES, practitioners can map information in MES onto the ADS.

Design Equations were used for anthropometric design of a truck cab. Functional Requirements were first formulated. Design Parameters were then selected with the purpose of creating an uncoupled design. Design equations were formulated, where couplings between functional requirements and design parameters were investigated. Ideally there should be a one-to-one correspondence between functional requirements and design parameters. This can be achieved if the design parameters are selected in a strategic fashion so that couplings are minimized. The nature of the design equations will inform the designer of the type of coupling: coupled design, semi-coupled design or uncoupled design. Design parameters that lead to couplings are replaced by other design parameters that create an uncoupled design. The design equations therefore serve to analyze a design, and by understanding the couplings it is feasible to synthesize a better solution. We call this method DESA — Design Equations for Systems Design. It has broad implications for design.

Most studies on product design have focused on customer needs concerning functionality and utility. Rarely has the issue of customer emotions been investigated. Traditional cognitive approaches to product usability tend to underestimate the importance of customer emotions in design. Not surprisingly, the success of a product in the marketplace may be determined by its aesthetic appeal, the pleasure it creates, and the satisfaction it brings to the customer. Emotions influence how a customer interacts with the product. In the interaction, feelings accompany thinking. This article characterizes customer emotions and issues concerning their measurement. The literature has documented various methods for measuring and evaluating emotions, but only a few are relevant to product development. In this article, we illustrate subjective and objective methods that are commonly used in evaluating customer emotion to an artifact. While evidence suggests that emotion matters in determining the sales potential of products, the measurement of emotions may be difficult because of the indefinite relationship between an emotion and its behavioral expression. Culture has a strong effect on what customers look for in a product and how they interpret the product and its interfaces. We present a framework for analysis of customer emotions in relation to the designer’s environment. The framework is part of the product development life cycle, where customer needs and measured emotions are input early in the design process. The goal is to achieve a pleasurable and satisfying product.

This research explores the use of axiomatic design principles for analyzing coupling in human-machine systems. Coupling is a generic dimension of complexity, and coupling between user goals and user actions widens the gulf of execution. An analytical framework was constructed based on the concept of design domains. This framework formalizes the interaction between user goals, functional requirements, design parameters, and user actions. User goals and user actions describe tasks from a user's perspective. Functional requirements and design parameters describe system's structure from an engineering perspective. Design equations are used to as a notation for representing the interaction between these attributes, and as a tool for characterizing the degree of coupling. By providing a general and rational criterion, the proposed method is conceived as an analytical tool that design engineers can use for early evaluations.

This paper summarizes underlying theories from psychology that can contribute to the further development of the science of Hedonomics. The methodology in this emerging area is immature. Although we do not know fully understand the mechanics of He-donomics, we can measure emotions and apply the results in designing artifacts which hopefully will be affective and pleasurable. A model for design of pleasurable artifacts and tasks is presented.

The study used Projective Testing techniques to identify the needs of elderly users of various features in a smart home, including: telemedicine, customized computer, teleconferencing, teleshopping, event reminder, robot helper and smart pets. Drawings of home scenarios and a male user were shown to 18 elderly users of service apartments. Altogether 192 different needs were identified. These were sorted by HF experts into broad categories. Nine categories of needs emerged relating to: Health, Communication, Security, Mobility, Mental Activities, Physical Activities, Prestige, Independence and Dependence. These were compared to needs structures documented in the literature including: Transcendent, Optimization, Anti-ageing, Dependency Avoidance and Nurturance Seeking. These needs structures are expressed at a higher level of abstraction then our nine categories. In addition the nine needs categories identified in this study can be used as a point of departure for the design of a smart home.

Mass customization aims at satisfying individual customers, who are given an opportunity to select among several design features and order a unique product. In this experiment four web sites were created where users could specify the design features of a mobile phone. Two web sites were designed to build a mobile phone from scratch: One used a rating agent of design features, and the other had no rating agent. The two other web sites used an existing phone as a basis for design - one used a rating agent and the other had no rating agent. Forty test persons participated in an experiment to evaluate the web sites. Half of them were male and half of them female. The use of the rating agent produced more consistent and better design. The model for mass customization — whether building from scratch or modifying an existing phone — had no impact on the results. There were no differences in results between males and females.

This paper presents a formal method for usability analysis. It can be used for design of products as well software. Functional requirements (FRs) and design parameters (DPs) are first stated. Then a conversion matrix from FRs to DPs is constructed. This matrix, referred to as “Design Matrix” is analyzed with respect to its properties. A diagonal matrix will have few usability problems, whereas a coupled design will have many usability problems. Usability may hence be improved by decoupling the matrix. This can be done by: 1. Selecting other design parameters. 2. “Reengineering” the problem by selecting other functional requirements and 3. Making sure that only one abstraction level at a time is addressed during the design of the interface.

Finke's notion of preinventive structure in top-down design was tested using Suh's methodology for top-down axiomatic design. Two groups of experimental subjects designed controls and displays for an “intelligent” automobile. One group was instructed to consider functional requirements and derive their design in a top-down fashion. The other group was also informed about functional requirements, and design methodology but was not instructed to use any particular design procedure. The first group produced significantly better designs than the second group - they were more creative, more practical and fulfilled design goals better. Top-down reasoning with careful consideration of functional requirements generated better design. Suh's methodology has an extra bonus, which is in agreement with Finke; it forces a slow deliberation of design features.

People have affective reactions towards tasks, artifacts and interfaces. Theses reactions are caused by design features, which operate either through their perceptual attributes or from a sense of control in handling or from past experience. A systematic framework is proposed to conceptualize Affective Human Factors Design or as we call it - Hedonomics. The word Hedonomics is derived from the Greek “hedone” (pleasure – akin to sweet) and “nomos” (laws, principles). The focus is on pleasurable design of artifacts and tasks. Several existing theories may combine to support this framework, including: Human Information Processing, Activity Theory, and Theory of Flow. There are also analogies with past theories – in human need structure, motivation, job satisfaction, and stress research. The problem is now to conceptualize Hedonomics, to propose theories that can be used for design, and appropriate measurement tools. It is particularly important to develop tools to measure and predict affective design – and on an individual level predict user and customer needs for affect.

A systems model is used to illustrate the information flow between three subsystems in e-commerce: Store Environment, Customer, and Web Technology. In the process of purchase, a customer makes several decisions: to enter the store, to navigate, to purchase, and to pay. This artificial environment must be designed so that it can support customer decision-making. At the same time it must be pleasing and fun, and create a task with natural flow. The paper summarizes existing research and theories, and provides a few suggestions for future research.

This paper outlines the use of cognitive walkthrough for design and evaluation of an airplane cockpit. Designers need methods that can help predicting the outcome of proposed design features at an early stage of design and later during design alterations and updating. Since the design of a military aircraft may take 10-15 years, it is important to develop evaluation methods other than user testing. This report proposes the use of Cognitive Walkthrough (CW) to evaluate the Pilot-Cockpit Interface.

This paper introduces Axiomatic Design (AD) for top-down design of ergonomics systems. The Independence Axiom was used to demonstrate how design activity can be structured to simplify the design process. Functional Requirements (FR) are first specified at a high level of abstraction and corresponding high level Design Parameters (DP) are generated. Then lower level Functional requirements are generated, and corresponding lower level Design Parameters. To avoid iterative improvements of design, one should strive for a decoupled design. This means that for each Functional Requirement there should ideally be only one Design Parameter - and this design parameter should satisfy only one functional requirement. Two case studies were used to illustrate the AD procedure.

Small batch manufacturing typical of modern production systems requires improved visual inspection of novel products for many defect types. Because each item inspected is different, inspectors cannot develop product-specific techniques to speed performance using traditional direct viewing. We performed two experiments to evaluate the use of binocular rivalry as an alternative to direct viewing in stereo viewing of product pairs. Both free stereo viewing (i.e., with no optical aids) and viewing with a stereoscope led to large increases in performance of searching for targets in extended symbol arrays. For increased target set size there was an expected increase in inspection time for direct viewing but no corresponding increase for stereo viewing. Binocular rivalry appears to be a promising, though difficult, tool for facilitating visual inspection, as it both improves performance and appears to allow parallel processing of targets.

This study investigated the hypothesis that excessive physical comfort adversely affects mood states generally associated with superior mental performance. This hypothesis was explored in an experiment on office chairs. Eleven subjects evaluated three chair designs varying in comfort for eight hours on three separate days. The subjects performed a word processing task from 8 a.m. to 5 p.m. Comfort and discomfort were evaluated using the Shackel Scale and the Corlett Scale. Mood states were evaluated using a mood profile survey. The results suggested that the most comfortable chair induced mental mood states such as boredom, tiredness, sluggishness, (lack of) alertness and (less) energetic feeling which are generally associated with lower mental performance. It is suggested that for knowledge workers, comfort and performance may follow an inverted-U relationship. Too little or too much comfort may be detrimental to both physical and mental performance. This moderation on the view of comfort has considerable implications for the design of the modern workplace, particularly for office furniture. Its relevance to the sedentary work-styles prevalent in the United States is also discussed.

We conducted a classification analysis to identify factors associated with sitting comfort and discomfort. The objective was to investigate the possible multidimensional nature of comfort and discomfort. Descriptors of feelings of comfort and discomfort were solicited from office workers and validated in a questionnaire study. From this study, 43 descriptors emerged. The 42 participants rated the similarity of all 903 pairs of descriptors, and we subjected the resulting similarity matrix to multidimensional scaling, factor analysis, and cluster analysis. Two main factors emerged, which were interpreted as comfort and discomfort. Based on these findings, we postulate a hypothetical model for perception of comfort and discomfort. Comfort and discomfort need to be treated as different and complementary entities in ergonomic investigations.

A controlled field study was conducted to measure spinal shrinkage in office employees with sit-stand type workstations. Thirteen office employees, ten healthy and three with spinal disorders participated in the study. Using a stadiometer, changes in stature were measured at 8 a.m., 12 p.m., 1 p.m. and 5 p.m. From 12 p.m. to 1 p.m., all subjects sat for 40 minutes and walked for 20 minutes. Out of the ten healthy employees, six were instructed to stand for 30 minutes four times during the day. The remaining four subjects stood eight times 15 minutes each. Office workers who stood in 30 minute sessions experienced significantly less shrinkage than those who stood in 15 minute sessions. Office workers with spinal disorders also stood eight times 15 minutes each. They incurred a greater variability in the shrinkage pattern.

Thirty eight subjects participated in a study to evaluate preferred (comfortable) settings of adjustable chairs including seat height, seat pan angle and back rest angle. An instrumented hydraulic chair was used. The chair was set in one of eight possible extreme position, with each chair parameters at either a high or low value. Subjects then adjusted all three parameters. A full factorial experiment with 5 factors was performed: the three adjustability parameters set at either high or low, gender and replication. Three independent variables were used: preferred seat height, back rest angle and seat pan angle. For the three ANOVA's and across two experiments, 33 statistically significant results were obtained. The initial setting of a parameter (high or low) consistently produced a significant effect in the choice of the preferred setting for the same parameter. The findings imply that there is a range of acceptable settings: for seat height about 1 in. and for the back support angle and seat back angle about 4 degrees.

Using Maynard's description, an assembly task may be divided into various task elements like reach, select, grasp, move, position, and assemble (Ghosh and Helander, 1985). Activities such as “reach” and “move” are governed by biomechanics of body motions including human factors principles such as Fitt's law. This research investigated the effect of the structure of the assembly and the type of assembly instructions on performance. Two different product structures were used — a vertical assembly, built bottom-up, and a hierarchical assembly, consisting of several subassemblies. Two different instruction strategies were used: 1. top-down sequential instructions 2. No Instructions. The research showed that vertical products were easier for manual assembly, when provided with instructions. For all three dependant measures, there was no interaction effect between the two factors, namely product structure and instructions. Average time for completion was significant for both factors i.e. product structures ( F(1, 20) = 4.417, p < 0.0485) and instructions ( F(1, 20) = 5.886, p < 0.0248). However, time for learning was significant only for product structure factor ( F(1, 20) = 5.239, p < 0.033). Also, trials to learn was significant only for product structure factor ( F(1, 20) = 4.449, p < 0.047).

Human factors and ergonomics are quickly evolving and new areas of application are introduced around the world. In industrially developing countries (IDC's) the emphasis was in the past on physical workload. With the introduction of computers there has been a sudden shift in interest, and the problems of HCI and usability of complex systems have become paramount. This panel will discuss the changes in ergonomics, and how they are affected by changing work ethics, and changing needs of the society. It is argued that ergonomics is well suited to adapt to new goals and respond to difficult challenges. As a result we need to prepare ourselves in education and training to meet the new demands in ergonomics. Can certification programs for ergonomists be flexible enough to consider current needs, and can teaching programs incorporate new knowledge? What are some effects on the type of research that will be conducted in the future. How will ergonomics professionals deal with design and applications. These are some of the issues that will be discussed in this panel session.

In order to facilitate automatic assembly by robots, it is necessary to reduce the number of parts to be assembled. A recent study at IBM analyzed the assembly paper feeding device for a copying machine that was composed of 27 different parts. This was too much for an automatic assembly procedure. After redesigning the product, the number of parts was reduced to 14, 13 of which could be assembled by a robot. The remaining part had to be inserted by a human operator. Such redesign usually incorporates several principles which facilitate robotic assembly. For example, the use of symmetrical parts so that the robot does not have to turn parts around, the use of notches to guide the insertion of parts, and a reduction in the number of screws used. The surprising outcome was that although the product could now be assembled by a robot, manual assembly had also been considerably simplified. In fact, it was so easy that it didn't pay off to use a robot. This redesigned part is still assembled by human operators. Obviously, changes that make the process easier for a robot might also simplify manual assembly. It seems ironic that we know more about the needs of robots than humans. This paper will discuss two examples of product design for assembly by robots.

This report summarizes a review of human factors research on VDTs. The main areas of interest included workstation design, character and display design, work organization, lighting and reflectance, and visual discomfort. Experimental studies were critically reviewed for the soundness of their methodologies and conclusions. In many of the studies, there were problems of interpreting the results due to careless oversights on the part of the researcher(s), for example, the inappropriate selection of subjects or lack of a control group. Due to different subject selection strategies and different VDT parameters, the integration and generalization of research results is difficult.

An increasing number of studies indicate that robots are the most hazardous equipment in industry. The very virtue that makes them attractive for industrial work, the programmable arm, is the cause of accidents since the arm motion is often difficult to perceive. The present paper presents a model of human reaction time and emergency behavior. The total reaction time is the sum of three elements: perception, decision making and motor response. Each of these three elements are modeled using concepts such as perceptual discriminability and single detection theory. Finally the results of an experiment is presented where the human reaction time is modeled as a function of robot arm speed.

Ten ergonomic office chairs, chosen from a sample of eighty-four, were evaluated in an ergonomic field study. Twenty office employees used each of the chairs for one day. The chairs were evaluated using four different methods: a chair feature evaluation checklist, a ranking procedure, general comfort and body part discomfort ratings. The first two methods produced positive results, including significant differences between the chairs. The chair users generally had difficulties in perceiving and expressing their comfort and discomfort associated with the latter two methods. The study identified several distinct features related to chair comfort, including the design of the seat pan, back rest, arm rests and ease of adjustability.

The Human Factors Society has decided to become increasingly involved in standards writing activities. This panel will give an overview of some of the most important human factors standards.

Machines used in low seem coal mines use manual hydraulic controls for most functions. This study addressed the optimum design of controls used for tramming (driving) a machine. For such tasks, the operator stoop walks behind the machine while pushing the controls. The study intended to determine at what height the controls should be located for a stoop walking task and what pressure would be considered maximum for control activation. Twelve male subjects were paid for their participation in the experiment. The following variables were manipulated within subjects: control pressure (10 and 20 pounds), seem height (40 and 50 in.), and control height (21, 26, 31, and 36 in.). The experiment was performed on a treadmill with a speed of 75 feet per minute. Each subject participated in 12 conditions, each lasting about 15 minutes for a total of three hours. Heartrate was measured continuously and sampled each minute. Oxygen uptake was analyzed for one subject. For the 40 in. seem height, the lowest control height (21 in.) and the highest control force (20 lbs) produced the lowest heartrate. For a 50 in. seem height, the lowest control height (26 in.) and the lowest control force (10 lbs) produced the lowest heartrate. The applicability of the results in terms of human factors design of mining machines is discussed.