Measuring and Facilitating Human-Computer Interaction

Ramsay, Judith Easton (1992) Measuring and Facilitating Human-Computer Interaction. PhD thesis, University of Glasgow.

Abstract

This thesis investigates two modes of facilitating repair in human-computer interaction (HCI). These two aspects of facilitation are mode of instruction, namely minimal manuals, and the use of emotional experience. In the first experiment, an additional 'from-scratch' design guideline was added to Carroll's (1990) guidelines for minimalist documentation in an attempt to formalise the source and type of content included in such manuals. A minimal manual was designed for e-mail using both his and the new guideline. The manual had approximately 13% of the pages of the commercial manual; it resulted in 30% faster learning and more effective use of the e-mail system overall, and significantly better performance on individual subtasks; including the recovery from errors. Significantly more users were satisfied with it than with the conventional manual. Carroll's general principles of manual design for minimal manuals were found to be a good basis for design, and it is suggested that his guidelines be combined with the proposed 'from scratch' method. Having formalised the last remaining guideline for design, it was then possible for the second experiment to consider the issue of the level of explanation of information required within the minimal manual. A second minimal manual was designed, which had a recipe-type format. This manual, once developed, was tested against the original manual designed in study one. No siginificant user performance differences were observed in a comparison test, however users reported overall that they subjectively preferred to use the original manual which promoted exploration and learning, rather than the recipe-style manual. Thus it would appear that making explicit instructional steps does not serve to further facilitate interaction. The latter part of the thesis attempted to further the understanding of human emotion within the HCI context. It was necessary to develop two new HCI measurement instruments to this end. The third experiment developed the first instrument. A checklist of cognitions and emotions, which when administered would give an indication of those cognitions and emotions. When administered, it would give an indication of those cognitions and emotions that are experienced when interacting with an interface. By a factor analytic procedure a set of 53 cognition and emotion statements was reduced to the manageable form of 10. This measurement instrument was subsequently employed in the fourth experiment. The aim of the fourth experiment was to measure human emotion during difficulties in interaction, and to discover whether the documented relation between emotion and facial expression holds true in the HCI context. If so, then recommendations could be made for the use of facial expression as a form of emotion feedback into computer systems, in the form of interface agents. 30 naive computer users completed a computer graphics task. During the task, their facial expressions were recorded by video, and their cognitions and state of progress were recorded, to see if there existed any relation between different types of episodes (getting stuck, making progress, and neutral episodes where nothing good and nothing bad is happening) and facial expression and emotion. The analysis of the users' facial expressions was conducted using an abridged version of Ekman and Friesen's (1978) Facial Action Coding System (the second instrument that had to be introduced by this thesis to the field). The occurrence and extremity of cognition and emotion experienced when users were stuck was measured by the checklist of cognitions previously designed and developed for this purpose. The facial expressions which were evident during episodes of progress, episodes of getting stuck, and neutral episodes were not significantly different from one another, despite there being a lot of emotional activity occuring. The thesis makes the following recommendations: (a) Firstly, that the design of user documentation should involve a synthesis of minimalist guidelines and a 'from-scratch' methodology. (b) Secondly, that cognitive coping techniques be developed alongside other methods of user assistance. (c) That it is necessary to not only develop new measurement instruments for HCI, but also to develop a framework to guide the optimal choice and combination of instruments employed.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Adviser: K Oatley
• Keywords: Experimental psychology
• Date of Award: 1992
• Depositing User: Enlighten Team
• Unique ID: glathesis:1992-75438
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 19 Nov 2019 20:07
• Last Modified: 19 Nov 2019 20:07
• URI: https://theses.gla.ac.uk/id/eprint/75438