Information seeking on websites has become more and more challenging for people who are blind. Screen readers attempt to make that task easier by introducing new commands and functionality but there still exists a fundamental gap in the difficulty for people who are blind compared to their sighted counterparts. This research identifies the strategies used by people who are blind and sighted to create an information seeking model, develop the Keywords Expected for Your Search (KEYS) conceptual model for assisting task performance and evaluate a KEYS prototype of this concept for information search and retrieval.
This information seeking model was developed by conducting an initial experiment and then validating the model using data from a similar study. The model defines information searching strategies used for both participants that were sighted and blind. Primary information search strategies for the web were identified. The model demonstrated the primary search strategies in the model were keyword based and are the foundation for information seeking on websites for both groups of participants.
The KEYS conceptual model for assisting task performance was developed. It includes rules and a keyword library to support information searching for users who are blind. A KEYS prototype was developed to test the conceptual model. The prototype was implemented by controlling the results of the list of links and virtual find commands, which are two of the more commonly used primary search strategies for users who are blind. The evaluation sought to determine the aids' impact on workload, number of commands used, search success, path direction and time to complete a task. Comparisons of the participants who were blind with and without the KEYS prototype were made as well as comparisons to sighted participants. Age group comparisons and age of blindness onset were also compared.
The results of the experiment demonstrated that the KEYS prototype significantly improved information searching for users who are blind by lowering all measured variables.
A primary contribution of this research was to demonstrate that providing keyword support improves information searches. Implementing the KEYS can produce a significant difference in how users who are blind search for information on webpages.