Research Method

---

Studying inter-organizational-trust in B2B e-commerce participation phenomenon requires the context of real organizations using e-commerce systems or operating in an e-commerce environment (Yin, 1994). A case study approach was seen as appropriate, particularly in an area where few previous studies have been conducted (Hart & Saunders, 1997; Sako, 1998). Organizations that participated in this study included a public sector organization involved in customs clearance, their Internet service provider, a customs agent (broker), and an importer.

Entry into the case sites was obtained by making initial telephone calls to key representatives in the e-commerce organizations. A brief description and purpose of the study was discussed over the telephone before requesting them to participate. The telephone conversation was followed by an email including an attached file describing the purpose of the study that was sent to the respondents. Most of the participants did indicate their interest to participate in the study and did request a case study report containing aggregate findings of all cases to be given to them at the end of the study. Once confirmation of their willingness to participate in the study was received, appointment dates for interview sessions were arranged at a convenient time suitable for the interviewees. Subjects were requested to answer focused questions, as well as open-ended questions via a semi-structured questionnaire in four (two-hour) session interviews. The interview sessions were recorded, and a draft report of their responses was sent back for verification and confirmation of their respondents. A report of the final findings, together with a thank you note, was sent to each organization that participated in the research.

Case Study Design - Data Collection and Analysis

According to Yin (1994), the following components make up the research design.

• A case study's questions, 'how' and 'why' types of questions, were considered in this study. How and why does inter-organizational-trust influence the perception of e-commerce benefits and risks, thus leading to the extent of its participation?

• The research propositions shown in Figure 1 help focus the study. Although the concept of trading partner trust is new to this study, evidence from the marketing and management literature clearly indicates that high levels of trust in business relationships can lead to low risks and controls (Cummings & Bromiley, 1996; Doney & Cannon, 1997; Gulati, 1995; Ganesan, 1994).

• Unit of analysis. This study identifies two units of analysis. The primary unit of analysis is the participants (that is, the e-commerce coordinators and managers involved in e-commerce participation). The next level of analysis is the inter-organizational-trust dyad, involving trading partners from organizations interacting with each other. Hence, even though only two organizations are involved in an inter-organizational-dyad, the relationship between the trading partners could be one-to-one, one-to-many, many-to-one or many-to-many relationships. The four case studies then formed inter-organizational-dyads shown (as oval shapes A, B, and C); that is, between NZ Customs and their ISP (A), ISP and customs broker (B), and the customs broker and importer (C). Furthermore, e-commerce applications, existing documents and standards were examined as units of analysis within each case. For example, trading partner agreements, organization charts, web sites, and internal security policies gave evidence of the organization's best business practices and background information of the organization and its products.

• The logic of linking the data to the propositions. Multiple case studies paved the way for both qualitative and quantitative analysis of the data collected. One method used to link the data and propositions of this study was via pattern matching across cases. The criteria for interpreting the findings were directly related to ways of linking data and propositions of study. The criteria applied included pattern matching and explanation building (as in narrative descriptions and causal explanations), thereby leading to analytic generalizations that formed a story line and predictions to be made. The first step involved identifying recurring patterns and themes from the data collected during the interviews. Then similarities and differences between the participants were identified, which was followed by a cross case analysis of the cases that paved the way for generalizations to be made.

---