eprintid: 9834
rev_number: 4
userid: 6
dir: disk0/00/00/98/34
datestamp: 2011-11-10 20:06:35
lastmod: 2016-11-15 13:52:17
status_changed: 2011-11-10 20:06:35
type: thesis_degree
metadata_visibility: show
contact_email: jms158@pitt.edu
item_issues_count: 0
eprint_status: archive
creators_name: Smith, Jennifer Marie
creators_email: jms158@pitt.edu
creators_id: JMS158
title: Exploring Human Computer Interaction and its Implications on Modeling for Individuals with Disabilities
ispublished: unpub
divisions: sch_hrs_healthandrehabsciences
full_text_status: public
keywords: disability; eye tracking; HCI; models; word prediction
abstract: Computers provide an interface to the world for many individuals with disabilities and without effective computer access, quality of life may be severely diminished. As a result of this dependence, optimal human computer interaction (HCI) between a user and their computer is of paramount importance. Optimal HCI for individuals with disabilities relies on both the existence of products which provide the desired functionality and the selection of appropriate products and training methods for a given individual. From a product availability standpoint, optimal HCI often depends on modeling techniques used during the development process to evaluate a design, assess usability and predict performance. Computer access evaluations are often too brief in duration and depend on the products present at the site of the evaluation. Models could assist clinicians in dealing with the problems of limited time with clients, limited products for the client to trial, and the seemingly unlimited system configurations available with many potential solutions. Current HCI modeling techniques have been developed and applied to the performance of able-bodied individuals. Research concerning modeling performance for individuals with disabilities has been limited. This study explores HCI as it applies to both able-bodied and individuals with disabilities. Eleven participants (5 able-bodied / 6 with disabilities) were recruited and asked to transcribe sentences presented by a text entry interface supporting word prediction with the use of an on-screen keyboard while time stamped keystroke and eye fixation data was collected. Data was examined to identify sequences of behavior, performance changes based on experience, and performance differences between able-bodied and participants with disabilities. The feasibility of creating models based on the collected data was explored. A modeling technique must support selection from multiple sequences of behavior to perform a particular type of action and variation in execution time for primitive operations in addition to handling errors. The primary contributions made by this study were knowledge gained relative to the design of the test bench and experimental protocol.
date: 2007-12-20
date_type: completed
institution: University of Pittsburgh
refereed: TRUE
etdcommittee_type: committee_chair
etdcommittee_type: committee_member
etdcommittee_type: committee_member
etdcommittee_name: Simpson, Richard
etdcommittee_name: LoPresti, Edmund
etdcommittee_name: Little, Roger
etdcommittee_email: ris20@pitt.edu
etdcommittee_id: RIS20
etd_defense_date: 2007-11-20
etd_approval_date: 2007-12-20
etd_submission_date: 2007-11-28
etd_access_restriction: immediate
etd_patent_pending: FALSE
assigned_doi: doi:10.5195/pitt.etd.2011.9834
thesis_type: thesis
degree: MS
committee: Richard Simpson (ris20@pitt.edu) - Committee Chair
committee: Edmund LoPresti () - Committee Member
committee: Roger Little () - Committee Member
etdurn: etd-11282007-193101
other_id: http://etd.library.pitt.edu/ETD/available/etd-11282007-193101/
other_id: etd-11282007-193101
citation:   Smith, Jennifer Marie  (2007) Exploring Human Computer Interaction and its Implications on Modeling for Individuals with Disabilities.  Master's Thesis, University of Pittsburgh.    (Unpublished)  
document_url: http://d-scholarship-dev.library.pitt.edu/9834/1/smithjm_dec_2007.pdf