Bulletin, February/March 2007

What's New?  Selected Abstracts from JASIST

Authors who choose to do so prepare and submit these summaries to the editor of the Bulletin.

From JASIST v. 57 (8)
Stock, W. G. (2006). On relevance distributions, 1126-1129.

Study and Results: In a theoretical study we found (based on empirical data) that there are at least three possibilities of how documents are distributed by relevance: informetric (power law), inverse logistic and dichotomous.

What’s New? The nature of the right relevance distribution has some implications. First, the notion of relevance is crucial for all those things called relevance ranking. If we do not know how documents are generally distributed by relevance, it is not possible to create a reasonable algorithm of relevance ranking. Second, it is also crucial for users of search engines to decide how many of the top-ranked documents they should read. Third, if a retrieval system works with automated, blind relevance feedback, it is important to know how many documents of the first ranking procedure are to be considered into the second step and are to be analyzed. Fourth, if we want to model score distributions for combining the outputs of search engines for meta-search, we need to know the nature of these distributions. Fifth, for topic detection and tracking, i.e., finding similar documents (e.g., news) on the same topic, relevance models are useful. Finally, evaluation exercises in information retrieval (e.g., TReC) usually work with a dichotomous view of relevance. So it is possible that an information retrieval system whose scores approximate the right relevance distribution (whichever it is) better than other systems does not receive a better evaluation assessment.

Limitations: None supplied

From JASIST V. 57 (9)
Rantanen, E. M., Palmer, B. O., Wiegmann, D. A., & Musiorski, K. M. (2006). Five-dimensional taxonomy to relate human errors and technological interventions in a human factors literature database, 1221-1232.

Study and Results: Traditionally, human error in aviation accidents has been addressed by different technological fixes, either by attempting to reduce the occurrence of errors or by mitigating the negative consequences of errors. However, new technologies and system changes may introduce new error opportunities or induce different types of errors. Thorough understanding of the relationship between error classes and technology fixes is therefore crucial for the evaluation of intervention strategies and effective direction of resources to maximize the benefit to flight safety.

What’s New? This paper describes mapping of intervention technologies onto error categories, creation of a conceptual framework and identification of applicable taxonomies for each dimension of the framework and construction of a usable prototype database. The framework consists of a three-dimensional matrix with axes for the human operator, the task and the environment. Human errors and technologies will cohabit molecules in the matrix linking them together. The database allows for taxonomic development in all three areas pertaining to human performance by keeping the taxonomies dynamic.

Limitations: A database can be completely tested only after it has been sufficiently populated. As classification and entry of relevant research literature into the database must be done manually by human factors experts, the full functionality of the database is deferred into the future. A Web interface for distributed database management is proposed to address this problem.

From JASIST v. 57 (10)
Lin, J., Chan, H. C., and Wei, K. K., (2006). Understanding competing application usage with the theory of planned behavior, 1338-1349.

Study and Results: The effect of competitive systems has not been considered in current studies on user acceptance and usage of information systems. Competitive effects are analyzed through the development of a relative model of the theory of planned behavior (TPB). Each factor (i.e., attitude, subjective norm, perceived behavioral control, intention and usage) of the TPB is measured relatively, with the value for one system minus the value for the other, for each subject response. Three hundred users of two instant messaging systems (MSN Messenger and ICQ) participated in a survey. The relative model shows the important factors to consider if we want to enhance user intention and usage of a system, relative to the competitive system.

What’s New? This article shows the importance of studying the relative model for competitive systems. If a company wants to know which factors to enhance, the correct answers are given by the relative model. In contrast, a single-system study may provide incorrect answers.

Limitations: Subjects in the study are undergraduate students in Singapore and may not reflect other populations, such as people from other age groups or other cultures. The generalization of this study to other environments should be made with caution.

Kifle, M., Mbarika, V., & Datta, P. (2006). Telemedicine in Sub-Saharan Africa: The case of teleophthalmology and eye care in Ethiopia, 1382-1393.

Study and Results: Developing countries are plagued by a lack of health care facilities, capital and specialists to support ongoing health needs. This study poses the question: Is telemedicine a viable alternative for countries lacking a requisite health infrastructure? The issue is examined in the context of implementing teleophthalmology solutions in Ethiopia. The findings suggest that once challenges such as lack of standards and accountability are mitigated and the workflow for the delivery of telemedicine is re-engineered, Ethiopia can gain substantial economic and social benefits.

What’s New? To date, implementation of telemedicine has remained mired in anecdotes. In contrast, this paper is grounded in case studies that tie telemedicine infrastructure challenges to potential benefits. A framework is developed as a template for review and use among practitioners interested in unraveling the process and outcome of telemedicine implementation in Sub-Saharan Africa. 

Limitations: The limitation of this paper is its single sample frame. It looks at an instance of telemedicine in the context of Ethiopia: an instance of a developing country. Spurious variables such as culture could have different influences across countries, requiring small changes in our framework. While limited in its contextual scope, we believe it provides an understanding of the conditions surrounding implementing telemedicine solutions in developing countries.