Preliminary findings on searcher performance and perceptions of performance in a hypertext bibliographic retrieval system
✍ Scribed by Wolfram, Dietmar ;Dimitroff, Alexandra
- Publisher
- John Wiley and Sons
- Year
- 1997
- Tongue
- English
- Weight
- 41 KB
- Volume
- 48
- Category
- Article
- ISSN
- 0002-8231
No coin nor oath required. For personal study only.
✦ Synopsis
Introduction
hypertext-based card catalog without prior instruction and that they searched as effectively when using a Boolean-Hypertext is a relatively new mechanism by which based catalog. bibliographic searching may be performed. Searcher con-Meadow, Wang, and Yuan (1995) examined perforfidence in using the system and in retrieving relevant mance and attitudes of process specialists and subject items may depend on several factors, such as the time domain specialists searching procedural and nonprocethey have taken or the number of record pages they have dural interfaces. They found that the interface affected visited in the database. Confidence in their performance performance in both groups. User attitudes differed by versus their actual performance may serve as an indicator group, too, with domain specialists more willing to read or ''reality check'' of the system's utility. This article records carefully but less willing than process specialists presents an initial exploration of the relationship between to learn more about manipulating the system. searcher perceptions of their performance and actual Although evaluation of hypertext system usage has search performance using two variants of a hypertextbeen studied, very little research exists that examines the based bibliographic retrieval system. relationship of searcher performance and perception of A number of studies have appeared that examine performance, particularly for hypertext-based information searcher behavior in full-text hypertext systems. Tagueretrieval systems for bibliographic data. The authors pres-Sutcliffe and Toms (1995) analyzed the search processes ent a preliminary investigation examining this topic. of novice searchers and found that searchers were more successful when they used the system's indexes rather Study Description than search commands, that repeated internal browsing negatively affected recall, that content summaries did not
The present research employed a prototype hypertext assist searchers in distinguishing between relevant and bibliographic retrieval system called HyperLynx. Hypernonrelevant text, and that available on-screen help was Lynx was designed to study hypertext system design and inadequate.
usage for highly structured information, specifically, bib-Marchionini et al. ( 1990) examined the effects of liographic records (Dimitroff & Wolfram, 1993, 1995). search and subject expertise on full-text hypertext-based Eighty-three subjects, recruited from the School of Lisearching. They found that search specialists exhibited a brary and Information Science and the Golda Meir Limore varied approach to searching than the subject spebrary at the University of Wisconsin-Milwaukee, were cialists. Both subject and search experts outperformed categorized as novice or expert searchers based on selfnovices.
rankings. They were then randomly assigned to either the In one of the few user studies that utilized bibliobasic or enhanced HyperLynx system. graphic records, Borgman et al. (1990Borgman et al. ( , 1991) )
report on
Each subject was asked to perform several searches a study that examined children's use of a hypertext-based after a brief orientation to the HyperLynx system. For card catalog. They found that children could search a this study, data results from two questions, entailing a keyword-based search with a small number of relevant records (8), referred to as Q Small , and another entailing a