摘要
在信息检索系统评估和信息搜寻行为研究中,实验法是常用的研究方法,而实验中的任务设计是实验成功的关键环节。设计模拟仿真工作任务情境是近年来常用的任务设计方式,而它能否替代用户的真实工作任务在很大程度上影响着实验研究结果的科学性。为验证此方法,招募42位本科生参加实验,要求其为完成一项真实工作任务和一项仿真工作任务,在CNKI中展开搜索。通过分析实验参加者在实施两项搜索任务中的交互信息检索行为和检索绩效,发现仿真工作任务与真实工作任务在属性上尽管存在某些显著差异,但并未对用户交互信息搜索行为及绩效产生显著影响。因而,仿真工作任务可以替代真实工作任务用于交互信息检索实验研究,但需要进行合理设计。
Experiment is a widely used research method in information retrieval systems evaluation and information seeking behavior research. Task design is critical in the experiment because it significantly affects research results. Simu- lated work task situations are used frequently in experiments in recent years. However, whether it could take place of us- ers＇ real work tasks affects the results of experiments to a great degree. This study conducts an experiment which recruits 42 undergraduate students. A simulated work task situation is designed and assigned to the participants to search in CNKI, who were asked to bring a real work task situation to the experiment. The paper analyzes characteristics of the two tasks in interactive information retrieval behavior and retrieval performance. Some significant differences in properties are recog- nized, but there are no significant effect found in users＇ behaviors and interactive performance when they engaged in two tasks. The paper concludes that simulated work tasks can replace real work tasks in experiment, but it needs to be careful- ly designed. Some related suggestions are discussed.
出处
《图书情报工作》
CSSCI
北大核心
2014年第16期5-12,共8页
Library and Information Service
基金
国家社会科学基金项目"基于用户多维交互行为的数字图书馆可持续发展评估模型与实证研究"(项目编号:11BTQ009)研究成果之一
关键词
仿真工作任务
真实工作任务
实验法交互信息检索
信息检索系统评价
simulated work task real work task experiment method interactive information search informationsystems evaluation
