Source
Automatically imported from: http://commons.somewhere.com:80/rre/1997/Information.Systems.Rese.html
Content
This web service brought to you by Somewhere.Com, LLC.
Information Systems Research
``` ---
This message was forwarded through the Red Rock Eater News Service (RRE). Send any replies to the original author, listed in the From: field below. You are welcome to send the message along to others but please do not use the "redirect" command. For information on RRE, including instructions for (un)subscribing, send an empty message to rre-help@weber.ucsd.edu
---
Date: Sun, 23 Mar 1997 14:53:35 -0800
From: "John L. King"
INFORMATION SYSTEMS RESEARCH Volume 8, Number 1 Contents and Editorial Notes
CONTENTS:
Strategic Choices in the Development of Inter-Organizational Information Systems by Vivek Choudhury
Why is Programming (Sometimes) so Difficult? Programming as Scientific Discovery in Multiple Problem Spaces by Jaewoo Kim and F. Javier Lerch
Redesigning Case Retrieval to Reduce Information Acquisition Costs by Vijay S. Mookerjee and Michael V. Mannino
Examining the Emergence of Hybrid IS Governance Solutions: Evidence from a Single Case Site by Carol V. Brown
Interorganizational Cooperation: The Role of Information Technology by M Bensaou
Measuring Software Engineering Evolution: A Rasch Calibration by Sasa Deklava amd David Drehmer
EDITORIAL NOTES
With this issue ISR begins its eighth year of publication. The 1996 year was a good year for ISR. We launched the new ISR format, expanded the number of pages allocated to the journal, and published our first special issue. There were a couple of surprises, however. One was the decrease in individual subscriptions to ISR, due (we think) to the effects of the merger of ORSA and TIMS into INFORMS and attendant changes in "bundling" policies for the journals with membership. Although our institutional subscriptions rose significantly, the decrease in indivdual subscriptions coupled with the larger page budget and commensurate costs resulted in ISR running a modest deficit.
For 1997 we will attempt to increase circulation by both individuals and institutions. We also will cut back a bit on the size of the first two issues to save money until our revenues rise. We hope the subscription and revenue problems will be resolved by mid-year. Also in 1997 we will launch the second ISR special issue effot, and we will strive (as always) to improve turnaround time in review of manuscripts. Generally speaking, ISR does pretty well on the most important measure of turnaround, namely reducing the time between the first submission and final printing. With one exception, all the papers published in this issues completed this cycle in less than 24 months. Still, we will try to do better with turnaround time for reviews.
I am pleased to announce that Arie Segev of the University of California, Berkeley, and Christopher Westland of the Hong Kong University of Science and Technology have joined the ISR Editorial Board. Henk Sol of the Technical University of Delft in the Netherlands has joined the Advisory Board.
A usual, an interesting and diverse set of papers in included in this issue. Choudhury examines the evolving issue of inter-organizational informaiton systems from the standpoint of a firm making strategic decisions about IOS development. The objective is to assist with the questions of what types of IOS might be useful, and how those IOS might be developed. Three types of IOS are described: electronic monopolies, electronic dyads, and multilateral IOSs. A study of IOS development in the aircraft parts industry permits evaluation of IOS development strategy in light of two transaction characteristics, demand uncertainty and market variability. Nine propositions are developed based on the empirical findings. In general, there is a close relationship between the character of IOS a firm develops and the relations the firm maintains with its trading partners, although it is suggested that adoption of a particular IOS could be used to alter relations with trading partners. Moreover, the findings suggest that the conventional view that IOS benefits the seller by raising switching costs for the buyer is incomplete, and in some cases broadcast or electronic sale systems might offer competitive advantage to sellers. The paper concludes with suggestions for practice and for future research on the subject.
Kim and Lerch explore the fact that programming in the development of effective software is often a difficult and frustrating process. They develop a conceptual view of programming as a three-space search problem, using rule spaces, instance spaces, and representation spaces. The central conjecture is that programmers must change the character of their representations when working in different problem spaces. In complex problems involving multiple problem spaces, this switching burden becomes significant. Three controlled experiments were conducted to investigate how programmers change representations, and the consequences of such changes. Three tasks were provided: writing of a new program (eight subjects), understanding of an existing program (24 subjects), and reuse of an old program for solving a new problem (nine subjects). The problems were all versions of the Tower of Hanoi isomorph. Protocol analysis was used for data collection. Results suggest that programming difficulty rises in direct relation to required representation change. Moreover, different programmers respond to changes in required representation in different ways. They draw a conceptual link between the response of programmers changing representations to the work of scientists engaged in the process of discovery, and suggest that programming could be a useful method for teaching general problem-solving skills if tasks are constructed to require representational changes.
Mookerjee and Mannino address the problem of finding extant cases in a data base given characteristics of a new case. This problem has been the focus of promising work in machine learning, but the cost-effectiveness of such approaches is hampered by separating the tasks of content formation and retrieval strategy formation. Four experiments are conducted to determine whether separated concept and retrieval strategy formation performs better than jointly optimized concept and retrieval strategy formation. The authors develop an algorithm to express the case retrieval loss criterion tha tselects attributes in ascending order of expected loss, and apply it to the separated strategy approach. They develop a cost-based variant of the ID3 algorithm, a decision tree in which attributes are selected using entropy reduction per unit of information acquisition cost, and apply it to the jointly optimized strategy approach. The results of the experiments suggest significant benefit from the jointly optimized strategy approach, although the data also suggest that information acquisition costs can be lessened even in the separated strategy approach if the appropriate cost considerations are applied.
Brown provides results from a detailed investigation of a large, multi-divisional firm moving from a uniform scheme of decentralized IS governance to a hybrid scheme involving centralization and decentralization of different IS functions. The study examines eight propositions regarding the utility of hybrid governance solutions based on a contextual framework of organizational action at the business unit level. Data on the propositions from both IS and non-IS personnel suggest that decentralized governance is appropriate in situations of organic decision-making, high business unit autonomy, a competitive strategy based on differentiation, and an unstable industry environment. However, even in cases where these conditions prevail, the recognition of deficiencies in IT capability will prompt adoption of alternative strategies that do not fit the normal pattern, particularly in business units with a predisposition toward organizational change. The concept of absorptive capacity is used to explain the behavior witnessed in the firm during the change to hybrid IS governance strategies. The pursuit of new tasks that require substantial specialized knowledge requires the presence in the business unit of substantial prior knowledge needed to assimilate the new knowledge. Lack of such prior knowledge makes a decentralized strategy difficult, and where re-centralization offers access to a wider pool of needed prior knowledge, can result in adoption of a hybrid solution. Implications for practice and for further work on governance solutions are discussed.
Bensaou investigates the use of information technology in maintenance of supplier relations between firms in the US and Japanese automobile industries. A theoretical framework based on multiple theories suggests that cooperation will be affected by three factors: the characteristics of the environment in which the cooperation takes place, the characteristics of the relationship between the two firms, and tthe extent of use of information technology across organizational boundaries. A set of 447 relationships in the purchasing and design domains were drawn from three US and 11 Japanese auto makers. Regression analysis conducted on five hypotheses revealed that all three factors affect the nature of cooperation, but there are some surprises. The climate in which the relationship exists is clearly significant, as is the nature of the specific relationship between the firms. The role of IT use differes significantly between the U.S. and Japanese firms, with substantially greater effect in Japan. Japanese auto makers appear to have exploited IT for the development of specialized, proprietary inter-organizational communications systems that strengthen already strong and long-lasting supplier-buyer relationships. In fact, Japanese auto makers have developed idiosyncratic systems that link together R&D laboratories, plants, and offices of the manufacturer directly to suppliers, distributors, and even to banks. U.S. firms, in contrast, appear to have applied IT less effectively through establishment of industry-wide technological platforms that do not capitalize upon or result in strong supplier-buyer relationships.
Dekleva and Drehmer use the Rasch calibration technique to provide a novel research report on the evolution of software engineering. The focus of the analysis is on the emergence of a set of metrics and models for software development capability embodied in the Software Engineering Institute's Capability Maturity Model (CMM). The CMM is essentially a device for assessing an organization's level of capability or maturity in development of complex software, using a set of criteria that the organization must meet. The more criteria met, the higher the level of maturity. The Rasch calibration method is used to determine whether necessary conditions for effective measurement have been reached, and can thus be used to evaluate and extend the data collection instrument used to determine maturity. Data were collected from 83 individuals in two cohorts of software maintenance experts, using a structured survey. Response analysis revealed that the two cohorts could be pooled, which was done for final analysis. The results of the Rasch calibration indicate that the CMM metrics do measure a general progression of firm capability toward maturity, but further show considerable clustering within levels and overlap across levels, suggesting that actual practice does not calibrate very well with the ideal of the CMM. A reconstruction of the stages implied by the data, using the CMM concepts, is provided by the authors. The authors conclude that new practices in software engineering appear to be implemented to correct problems in practices adopted earlier. Also, it appears that some practices are only beneficial when implemented in the context of other practices, suggesting the measurment of clustered, of complementary practices could provide a better indicator of maturity. ```
This web service brought to you by Somewhere.Com, LLC.