Participative Analysis of Systems Integration Opportunities

View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by AIS Electronic Library (AISeL) Association for Information Systems AIS Electronic Library (AISeL) AMCIS 1997 Proceedings Americas Conference on Information Systems (AMCIS) 8-15-1997 Participative Analysis of Systems Integration Opportunities Ann M. Hickey University of Arizona, ahickey@bpa.arizona.edu Douglas L. Dean University of Arizona, ddean@bpa.arizona.edu Douglas R. Vogel University of Arizona, vogel@bpa.arizona.edu Follow this and additional works at: http://aisel.aisnet.org/amcis1997 Recommended Citation Hickey, Ann M.; Dean, Douglas L.; and Vogel, Douglas R., "Participative Analysis of Systems Integration Opportunities" (1997). AMCIS 1997 Proceedings. 211. http://aisel.aisnet.org/amcis1997/211 This material is brought to you by the Americas Conference on Information Systems (AMCIS) at AIS Electronic Library (AISeL). It has been accepted for inclusion in AMCIS 1997 Proceedings by an authorized administrator of AIS Electronic Library (AISeL). For more information, please contact elibrary@aisnet.org. Participative Analysis of Systems Integration Opportunities Ann M. Hickey ahickey@bpa.arizona.edu Douglas L. Dean, Ph.D. ddean@bpa.arizona.edu Douglas R. Vogel, Ph.D. vogel@bpa.arizona.edu Center for the Management of Information University of Arizona, Tucson, Arizona 85721 Abstract In an effort to increase information sharing while simultaneously decreasing costs, many organizations are moving to integrated data and systems. However, researchers caution that the costs and benefits of integration must be carefully evaluated. This paper presents a participative integration analysis methodology for determining not only what "can" be integrated, but also what "should" be integrated. Results of the initial case study show that a small group can effectively decide what "should" be integrated and develop a proposed integration strategy. The results also highlighted that participants intuitively used business scenarios to identify integration opportunities and analyze the business impacts of integration. Therefore, the participative integration analysis methodology was updated to incorporate scenarios as the central evaluative construct. This methodology will result in recommendations for integrated systems and business processes. Introduction In today's increasingly competitive global business environment, organizations must devise cost-effective mechanisms for rapidly sharing information across organizational sub-units and with external business partners. From an information systems perspective, this problem has been traditionally addressed through systems and data integration, i.e., the development of enterprise information architectures, integrated systems, or corporate databases with common data definitions and formats. However, studies have shown that integration has often either failed or achieved limited success in many large organizations (e.g., see Goodhue, Kirsch, Quillard, & Wybo, 1992a). Goodhue et. al. (1992b) attribute these problems to: (1) implementation pitfalls, (2) shortcomings in the methodologies, and (3) integration costs which exceed its benefits. Implementation pitfalls such as the need for organizational support can be partially addressed through a participative analysis process which involves key organizational stakeholders. Shortcomings of data integration methodologies based on the need for human classification of complex data relationships can also be addressed through a participative analysis process involving users from the appropriate organizational sub-units. The need to compare the costs and benefits of data integration can be addressed through a participative integration impact analysis which determines whether those impacts are acceptable to the user from a business perspective and feasible from a systems perspective. The purpose of this research is to develop and evaluate a methodology which facilitates the participative evaluation of what organizational data "can" and "should" be integrated - a critical prerequisite to the development of integrated organization information systems. This research is a synthesis and extension of previous research in database schema integration, data integration impact analysis, and participative modeling. The next section includes a brief summary of this earlier research and a description of our model for participative integration analysis. Our research methodology is summarized in the third section with the results from the initial case study described in the fourth section. The paper concludes with a discussion of the implications of those results and a summary of the contributions made by the current research effort. Participative Integration Analysis Database research has clearly defined the many benefits of data integration including reduction in data redundancy and inconsistency as well as increased data sharing (Date, 1995). Based on these results, there has been an implicit assumption that organization data should be integrated whenever possible. Database schema integration research has focused on how to accomplish this integration (Batini, Lenzerini, & Navathe, 1986; Ram & Ramesh, 1996). A consistent requirement of schema integration methodologies is that users or database administrators (DBAs) must identify the relationships between the objects in the schemas being integrated (Spaccapietra & Parent, 1994). However, in large organizations, this task is often too complicated to be performed centrally by a few individuals (Spaccapietra & Parent, 1994). Users and DBAs from each of the business areas need an efficient mechanism for working together and sharing their domain knowledge to make these complex decisions on what data "can" be integrated. Another complicating factor in the data integration process raised by Goodhue et. al. (1992b) is whether the benefits of data integration always outweigh its costs. They recommend a careful comparison of the benefits of integrated, sharable data and systems versus the costs of decreased local autonomy and increased complexity of IS development before determining whether the data "should" be integrated (Goodhue et al., 1992b). Our proposed model of participative integration analysis addresses both these concerns. This approach is an extension of previous research on using electronic meeting systems to support group process and data modeling. The research has clearly shown the benefits of user involvement for developing complex models of a business area (Dean, Lee, Orwig, & Vogel, 1994-95; Dean, Lee, & Vogel, 1997). Participative modeling has been especially effective when input is required from large, heterogeneous, groups of users. Based on these results, we theorized that bringing together similar groups from the business areas to be integrated would be an effective approach for analyzing integration opportunities. Specifically, we theorized that a small group of users, analysts, and data modelers from those areas, using the database schemas and their knowledge of each business area, could effectively determine what data "could" and "should" be integrated. Research Methodology Our basic research methodology was an iterative process of developing our proposed model for participative integration analysis, evaluating it through exploratory case studies, and then refining the model based on the results of the case studies. This paper reports on our initial participative integration analysis model, the results of our first case study, and our proposed refinement of the model which will be evaluated in future case studies. The first case study explored the viability of our participate integration analysis model by evaluating whether a small group of users, analysts, and data modelers could work together to determine an efficient data and system integration strategy. The case study was used to evaluate whether the group could: (1) decide what data "could" and "should" be integrated, (2) develop a data and systems integration strategy, and (3) provide the information needed to gain approval and execute the integration strategy. The effectiveness of the proposed participative integration analysis approach was also evaluated. Results The initial case study centered around an integration meeting held to determine if two DOD Environmental Security business areas could and should be integrated. Attendees included two Hazardous Material and Air Emissions Control users, systems analysts working in the two areas, the Air system's data modeler, Environmental Security systems development managers, and the researchers. Hazardous Material and Air database models had been previously developed using the group data modeling tools described in (Dean et al., 1997). During the Air data modeling session, users also identified potential opportunities for integration. The integration meeting began with an overview of the two business areas and the existing data models to ensure a common understanding of both areas and to informally discuss possible integration. Building on that discussion and the integration ideas generated in the earlier Air data modeling session, the group used GroupSystems to identify and categorize the commonalties and uniqueness of the business areas. The group then analyzed each common category, reviewing entities and attributes to determine whether they could be integrated. To make this determination, the users would describe specific examples or scenarios of how and when they would use the data. The users continued to identify additional scenarios until the group felt they had sufficient knowledge to determine whether the items "could" be integrated. If the items could be integrated, the group then discussed how integration would change the way they did business and whether that change was acceptable. The group brainstormed and analyzed potential "integrated" business scenarios and system implementation alternatives to make this determination. If the changes were acceptable, the group determined that the data "should" be integrated. At the end of the session, the group decided that the amount of data which "should" be integrated indicated a need for development of an integrated system. They therefore discussed possible strategies for integration and prioritized integration based on both the benefit to the user and the ease of system implementation. Meeting participants were extremely satisfied with both the participative integration analysis approach and the proposed integration strategy. Since the meeting, the Air and Hazardous Material data modelers successfully integrated the database schemas and the integrated schemas are being phased into the current Hazardous Material system as proposed. Discussion These results showed that a small group of users, supported by analysts and a data modeler, can participatively determine what "can" and "should" be integrated. The success of this group validated our initial participative integration analysis model. More importantly, however, the results provided us with critical knowledge that will enable us to further enhance and structure the integration analysis process. Participants intuitively used business scenarios to describe their use of the data so they could determine whether data "could" be integrated. They also brainstormed potential integrated scenarios to determine the acceptability of the business changes resulting from integration and, therefore, whether the data "should" be integrated. Based on these results, we revised our participative integration analysis model to highlight the central nature of scenarios in identifying and analyzing integration opportunities. We are also exploring methodologies for eliciting business scenarios from groups of users which more completely describe each business area. Having a validated set of business scenarios will improve the integration analysis by relieving participants of the need to generate scenarios and ensuring that the group considers a comprehensive set of data uses and potential integration impacts before they recommend what "should" be integrated. Recommendations for "integrated" business processes can also be identified based on the integrated business scenario analysis. Conclusion This paper summarizes our initial efforts to respond to Goodhue et al.'s call for "methods that can help an organization determine which data should be integrated and which should not" (Goodhue et al., 1992b, p. 308). We have begun to meet this challenge with our proposed model of participative integration analysis which includes: (1) identification and analysis of integration opportunities using a participative process supported by an electronic meeting system, (2) evaluation of not only what "can" be integrated, but also what "should" be integrated, and (3) an assessment of how integration impacts business processes through evaluation of integrated business scenarios. References Batini, C., Lenzerini, M., & Navathe, S. B. (1986). A comparative analysis of methodologies for database schema integration. ACM Computing Surveys, 18(4), 323-364. Date, C. J. (1995). An Introduction to Database Systems. (6th ed.). Reading, MA: Addison-Wesley Publishing Company. Dean, D. L., Lee, J. D., Orwig, R. E., & Vogel, D. R. (1994-95). Technological support for group process modeling. Journal of Management Information Systems, 11(3), 43-64. Dean, D. L., Lee, J. D., & Vogel, D. R. (1997). Group tools and methods to support data model development, standardization, and review. In J. F. Nunamaker & R. H. Sprague (Eds.), Proceedings of the Thirtieth Annual Hawaii International Conference on the System Sciences (Vol. II, pp. 386-420). Los Alamitos, CA: IEEE Computer Society Press. Goodhue, D. L., Kirsch, L. J., Quillard, J. A., & Wybo, M. D. (1992a). Strategic data planning: Lessons learned from the field. MIS Quarterly, 16(1), 11-34. Goodhue, D. L., Wybo, M. D., & Kirsch, L. J. (1992b). The impact of data integration on the costs and benefits of information systems. MIS Quarterly, 16(3), 293-311. Ram, S., & Ramesh, V. (1996). Schema integration: Past, current and future. In A. Elmagarmid, A. Sheth, & M. Ruzinkiewicz (Eds.), Management of Heterogeneous and Automated Database Systems. Morgan Kaufman.Spaccapietra, S., & Parent, C. (1994). View integration: A step forward in solving structural conflicts. IEEE Transactions on Knowledge and Data Engineering, 6(2), 258-274.