A conceptual multi-model HCI model for the blind

Abstract

The ability for blind people to read and write Braille aids literacy development. A good level of literacy enables a person to function well in society in terms of employment, education and daily living. The learning of Braille has traditionally been done with hard copy Braille produced by manual and more recently electronic Braille writers and printers. Curtin University is developing an electronic Braille writer and the research on an interface for Braille keyboard devices, presented in this thesis, forms part of the Curtin University Brailler project.The Design Science approach was the research method chosen for this research because of the flexibility of the approach and because it focuses upon the building of artefacts and theory development. The small sample size meant that both individual interviews and a focus group were employed to gather relevant data from respondents. The literature review covers a variety of areas related to computer interfaces and Braille keyboard devices. A key finding is that the interaction paradigm for Braille keyboard devices needs to differ to interfaces for sighted individuals because of the audio, tactile and serial nature of the information gathering strategies employed by blind people as compared with the visual and spatial information gathering strategies employed by sighted individuals. In terms of usability attributes designed to evaluate the interface consistency was found to be a key factor because of its importance to learning and memory retention.However, two main functions carried out on a computer system are navigating and editing. Thus the model of interface for Braille keyboard devices presented in this thesis focuses upon navigation support and editing support.Feedback was sort from by interviews with individuals and a focus group. Individual interviews were conducted face to face and via the telephone and the focus group was conducted via Skype conference call to enable participants from all over the world to provide feedback on the model.The model was evaluated using usability attributes. Usability was important to the respondents, in particular consistency, learnability, simplicity and ease of use were important. The concept of rich navigation and infinitely definable key maps were understood by respondents and supported. Braille output is essential including the ability to show formatting information in Braille.The limitations of the research included the few respondents to the interviews and the choice to focus upon a theoretical model rather than implementing the model on an actual device. Future research opportunities include implementing the interface concepts from the model on to touch screen devices to aid further development of the interface and implementing the interface on a physical device such as the Curtin University Brailler.