Gerhard P. Deffner

Evaluation of Concurrent Thinking Aloud Using Eye-tracking Data

In a series of studies we addressed the two questions of: 1) Do verbalizations reflect concurrent thought, and 2) Does concurrent thinking aloud differ from normal thinking? The design of experimental tests was based upon Ericsson and Simons model of thinking aloud, incorporating variations of how information is represented in short term memory. Eye-movement recordings were used as a source of additional data, allowing us to go beyond a mere analysis of solution time and accuracy. Comparing verbalizations and eye-movement data, we arrived at a positive answer to the first of our initial questions. The second question was approached on several levels, always involving a comparison of ‘think-aloud groups’ with silent controls. We found no differences with respect to accuracy, but longer solution times in think-aloud groups. In a final experiment, the influence of thinking aloud on concurrent task performance could be narrowed down to an effect which only persists through the early stages of familiarization with tasks. We conclude that concurrent verbalization is a viable tool in the study of cognitive processes.

Eye Movement Recordings To Study Determinants Of Image Quality In New Display Technology

Abstract The paper describes and illustrates a paradigm for in-depth study of perceived quality of visually displayed images. The development of new, all-digital display technology offers unique possibilities for fine tuning the match between display devices and the human perceptual system to achieve superior image quality. Engineering efforts need to be prioritized, however, according to which image characteristics are the most important for the perception of quality. Detailed studies have frequently run into problems because subjective data have not been very reliable. The present approach uses eye-movement data recorded while subjects view and compare images on different displays. These recordings provide quantitative data, indicating which parts of the test images received how much of a viewer’s attention. In addition, these recordings are played back to provide cues for subsequent retrospective verbalizations, providing more data about a subject’s thought processes when judging perceived image quality.

Should we or shouldn't we use spoken commands in voice interfaces?

INTRODUCTION Letting users speak commands to the systems they are interacting with has been tried in various domains. Examples include a range of computer applications, as well as more narrowly defined domains such as aircraft systems. Using speech to control computers has never caught on in a big way, however, and in aircraft it has been limited to a few special applications. Why? Isn’t it appealing to be able to simply say “justify” when doing word processing (instead of having to give that command through pressing a command key or making a selection from a menu)? Doesn’t it make a lot of sense to let a pilot concentrate her/his attention on the main job of flying instead of fumbling for the radio and manually setting it to the new frequency?

User acceptance and preference for advanced voice services features and dialogue styles

The experiment evaluates several alternatives for the design of user dialogues of a telephone system which integrates advanced features to accelerate access to telephone services and also Spoken Speed Dial, Call Answering and Call Delivery. In the experiment, subjects placed phone calls and relayed pieces of information. The dialogue they then encountered presented them with experimental variations of dialogue styles and provided an opportunity to use the Spoken Speed Dial feature. Subjects made 16 call attempts in the first phase of the experiment, and their preferences for dialogue features were recorded between trial blocks and at the end of the experiment. In a second phase, subjects were the recipients of Call Delivery of various types. The results show clear preference for verbal prompts, the usability of Spoken Speed Dial, conflicting attitudes towards the extra step of recording the recipients name in Call Answering, and a preferred mode of Call Delivery.

Visualization of manufacturing process data in N-dimensional spaces: a reanalysis of the data

As process engineers are forced to better understand and control their manufacturing processes, the ability to identify the relationships between variables becomes important to the control of the processes. There are various methods available to facilitate the identification of quantitative relationships, but all are somewhat limited in their ability to characterize important relationships involving large numbers of variables and vast amounts of data. A previous study compared the ability of subjects to identify relationships on the basis of several different modes of displaying process data. The study also explored options for data visualization techniques that would aid the identification of complex relationships by combining data display capabilities of high resolution graphic work stations and the pattern recognition capabilities of humans. Results showed that subjects who could accurately use a three-dimensional display had faster response times than subjects who used a two-dimensional display. This paper discusses a more in depth analysis of the data from the previous study. Specifically, it examines the influences of data visualization style, perceptual complexity, and informational complexity on the users response times and accuracy. The reanalysis confirms and further explains earlier findings. That is, three-dimensional displays improve performance when information displayed is perceptually complex, whereas informational complexity is best displayed in two dimensions. The applicability of this research to process characterization, statistical analyses, and other software tools is discussed.

Verbal Protocols as a Research Tool in Human Factors: Panel Discussion

Verbal protocols have been used for many years in different research contexts, but there still is no clear consensus about the validity of the technique and methods for maximizing validity in an applied setting; how to standardize the collection and analysis of protocols; and last but certainly not least, whether the resulting data is worth the effort. This panel discussion is a companion to a symposium at this conference which presents empirical studies and human factors applications of verbal protocol techniques. The panel will focus in more depth on issues raised in that earlier session, with the goal of providing guidance for practical applications of the technique.

On the Use of a Primitive Formal Language and Ill-Defined Quantifiers in Knowledge Acquisition

This paper describes a tool for knowledge acquisition and its empirical evaluation. The main characteristics are the use of a highly restricted language allowing for the expression of rules by means of cards and their combination. In order to assess the feasibility of the approach, subjects were trained in the use of a computer simulation that only gave them very general, unspecific, feedback about the internal workings of the system. It was their newly gained knowledge about how to control the system that was probed in a subsequent knowledge acquisition step. We observed substantial accuracy and also demonstrated its validity in a comparison against prior performance.