Date of Award

December 2013

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Information Studies

First Advisor

Iris Xie

Committee Members

Dietmar Wolfram, Xiangming Mu, Pertti Vakkari, Diane Kelly

Keywords

Digital Library, Search Tactic, System Support, User Engagement

Abstract

This study aims to investigate users' search tactic application and system support in using digital libraries. A user study was conducted with sixty digital library users. The study was designed to answer three research questions: 1) How do users engage in a search process by applying different types of search tactics while conducting different search tasks?; 2) How does the system support users to apply different types of search tactics?; 3) How do users' search tactic application and system support for different types of search tactics affect search outputs? Sixty student subjects were recruited from different disciplines in a state research university. Multiple methods were employed to collect data, including questionnaires, transaction logs and think-aloud protocols. Subjects were asked to conduct three different types of search tasks, namely, known-item search, specific information search and exploratory search, using Library of Congress Digital Libraries. To explore users' search tactic patterns (RQ1), quantitative analysis was conducted, including descriptive statistics, kernel regression, transition analysis, and clustering analysis. Types of system support were explored by analyzing system features for search tactic application. In addition, users' perceived system support, difficulty, and satisfaction with search tactic application were measured using post-search questionnaires (RQ2). Finally, the study examined the causal relationships between search process and search outputs (RQ 3) based on multiple regression and structural equation modeling.

This study uncovers unique behavior of users' search tactic application and corresponding system support in the context of digital libraries. First, search tactic selections, changes, and transitions were explored in different task situations - known-item search, specific information search, and exploratory search. Search tactic application patterns differed by task type. In known-item search tasks, users preferred to apply search query creation and following search result evaluation tactics, but less query reformulation or iterative tactic loops were observed. In specific information search tasks, iterative search result evaluation strategies were dominantly used. In exploratory tasks, browsing tactics were frequently selected as well as search result evaluation tactics. Second, this study identified different types of system support for search tactic application. System support, difficulty, and satisfaction were measure in terms of search tactic application focusing on search process. Users perceived relatively high system support for accessing and browsing tactics while less support for query reformulation and item evaluation tactics. Third, the effects of search tactic selections and system support on search outputs were examined based on multiple regression. In known-item searches, frequencies of query creation and accessing forwarding tactics would positively affect search efficiency. In specific information searches, time spent on applying search result evaluation tactics would have a positive impact on success rate. In exploratory searches, browsing tactics turned out to be positively associated with aspectual recall and satisfaction with search results. Based on the findings, the author discussed unique patterns of users' search tactic application as well as system design implications in digital library environments.

Share

COinS