| Abstract: | The central research question of this thesis is: How to inform trustworthiness assessments of virtual project team members in the initial phase of collaboration?
There is common agreement that the availability of personal information and the possibility to interact informally at the start of a project accelerates the trust formation process. This goes for face-to-face as well as for virtual project teams. However, there is no shared understanding as to what information is critical for this acceleration and why it is so. Acceleration of the trust formation process is beneficial, as interpersonal trust is one of the key factors influencing performance in face-to-face as well as virtual teams. When little or no trust exists within a team, serious collaboration problems are bound to occur. Virtual project teams experience more problems with interpersonal trust formation than face-to-face teams. This is likely to be due to the diminished availability of information and its computer-mediated character. Once we know what information is important for trustworthiness assessments and why it is so, we could use it for the design of measures to accelerate the formation of interpersonal trust. To investigate the central research question we combined a theoretical (top-down) with a practical, design-oriented (bottom-up) research approach. We concluded our research with an evaluation.
1. Theoretical perspective
Chapter 2 reports the results of a literature study undertaken to gain insight in the interpersonal trust formation process, the various factors contributing to an interpersonal state and the differences between trust formation in face-to-face versus virtual project teams. It showed that perceived professional trustworthiness is an important determinant of the overall interpersonal trust state of a trustor (person who trusts someone else). It also indicated that virtual project teams lack the signs and signals needed to assess this professional trustworthiness. We therefore developed a first version of a schema for perceived professional trustworthiness, the TrustWorthinessANtecedent schema (TWAN), to provide insight in which antecedents trustors take into account while assessing the professional trustworthiness of a trustee (person who is trusted by a trustor). This schema consists of 23 hypothesized antecedents, which can be seen as general characteristics of trustworthy professionals. Examples of such antecedents are ‘communality’- the personal characteristics which the trustor has in common with the trustee; and ‘knowledge’- the extent to which a person recalls facts, concepts, principles and procedures within certain domains. Furthermore, we discovered that virtual project team members specifically lack cues for the assessment of professional trustworthiness of a trustee. Therefore, we introduced a profile template as a method to support trustworthiness assessments in virtual project teams. This was the starting point for the design-oriented research, described in Chapter 4 and Chapter 5, in which the perceived user value of a profile and the relation between specific information elements and professional trustworthiness antecedents was investigated. Starting from the TWAN schema, Chapter 3 examined on which antecedents professional trustworthiness assessments are based in practice. A large number of professionals with extensive collaboration experience were canvassed on their trustworthiness assessments of two colleagues in a non-hierarchical relationship. The questionnaire was, wherever possible, based on previously used measurement items. Firstly, the scales within the questionnaire were assessed on their reliability and validity. This led to an adaptation of the TWAN schema, corresponding scales, and the questionnaire. ‘Self-confidence’ did not measure perceived trustworthiness and ‘consistency’ could not be operationalised with sufficient quality. Subsequently, we tested to what degree the TWAN schema and a measure for Trust predisposition (a stable attitude towards trust-related information) could predict a general value for interpersonal trust (as measured with a widely accepted scale). Although all 21 remaining antecedents were used, seven antecedents proved to be most useful to assess interpersonal trust after extensive collaboration. These were the antecedents of ‘communality’, ‘sharing’, ‘responsibility’, ‘skills’, ‘persistence’, ‘caring’ and ‘discretion’. Trust predisposition turned out to be less predictive than expected on the basis of previous research results, which may be due to our measuring professional trust after extensive collaboration rather than in the initial contact phase.
2. Design-oriented perspective
The case study presented in Chapter 4 explored if and how virtual project team members value having a profile available to inform their trustworthiness assessments. The study also examined whether such a profile should contain static (fixed, not subject to updates) and/or dynamic (changeable, continuously updated) information elements. Information elements are ‘containers’ for information. Examples of information elements and corresponding information (between brackets) are ‘name’ (Klaas Visser), ‘occupation’ (plumber) and ‘testimonials (references from another person)’ (delivers neat work and keep one’s appointments). Results showed that a profile template with static information was especially useful to become acquainted with each other in the first two to three weeks. After this period impressions are mostly based on work-related performance, communication patterns and collaborative behaviour. If no profile was available in the initial phase, team members’ impressions were grounded in whatever information was available. This finding provides evidence for the fact that people construct an impression of each other’s trustworthiness irrespective of the amount and quality of the information that is available. When the profile was available on time, all members used it: they filled it in as well as read the descriptions of their team members. Team members were divided on whether to explicitly represent dynamic information in a profile (for example an indicator for the responsiveness on messages), as they derive this information themselves from context as well as interactions while collaborating. The case study showed that a profile template can support trustworthiness assessments in the initial phase of virtual team collaboration. However, it didn’t provide insight in what information elements in the profile were especially informative for these assessments. Therefore, the study described in Chapter 5 inventoried information elements perceived as important by designers as well as trustors. Trustors with virtual project team experience were here asked to rate the importance of information elements for their trustworthiness assessments as well as their practical relevance for collaboration. They rated a pre-defined list with information elements. This list was composed of information elements found in the profiles within 17 trust-requiring environments, thus representative for the designers’ perspective. We expected that personal, not-work related information would on average be rated high across the group. However, the commonly preferred information elements especially referred to trustworthy characteristics in a professional context, such as a trustee’s ability, motivation and availability. Several information elements that were made available by designers across all trust-requiring situations (e.g., name, photo) seem to be taken for granted by trustors, as they were not among the information elements ranked highest. Trustors labelled only a few dynamic information elements as important. Interestingly, both lists with information elements differed significantly, depending on the designers’ or trustors’ perspective. The elements identified as important for practical reasons overlapped to a large extent with the information elements identified by trustors as important for trustworthiness assessments. The information elements that were deemed to have practical relevance were almost all related to the availability of other people, the language used within the project, and communication methods employed to contact each other. Some elements were identified as only relevant for practical purposes, such as the ‘local time at the location of a team member’. The studies reported in Chapter 4 and Chapter 5 resulted in a subset of the original set of information elements; all the elements in the subset helped to facilitate initial trustworthiness assessments (Appendix J).
3. Evaluation
Chapter 6 focused on the question which information elements trustors value particularly and whether these elements do indeed reveal relevant cues for the trust warranting characteristics of a trustee. To that end, 226 trustors were asked to select the information elements that were most important for their trustworthiness assessments at the initial stage of collaboration. The trustors were canvassed on the inferences they could derive from these elements. Their rationales were then analysed using a coding scheme based on the TWAN schema to check whether information preferences were related to the professional trustworthiness antecedents or perhaps to other concepts. Respondents predominantly valued information elements that provided them with multiple cues to the trustworthiness of a trustee, possibly indicating an ‘information efficiency’ strategy of trustors. Examples of such information elements are ‘personality traits’, ‘age/date of birth’ and ‘work experience’. Information elements providing unique cues only could not be identified. Some additional, more personal information elements were identified as having value for professional trustworthiness assessments (for example ‘photo’, ‘interest/hobbies’ and family situation/marital status). The antecedents of professional trustworthiness that were more frequently mentioned across explanations were ‘competence’, ‘commitment’, ‘responsibility’, ‘availability’ and ‘communality’. Results also indicate that not all information preferences can be explained with the cognitive schema of trustworthiness alone. For example, the appreciated information element ‘photo’ seemed to be selected especially because it provides trustors with a certain intuitive ‘feeling’ for a trustee.
Conclusion
From the results, we may conclude that providing trustors in the initial phases of collaboration with information that matches their specific cognitive needs, as further detailed in the TWAN schema, helps to inform trustworthiness assessments. A profile template is fit to ensure the provision of this information. Contrary to our expectation, we found that antecedents receive different emphases depending on the collaborative stage. In the initial phase, the antecedents ‘competence’, ‘commitment’ and ‘availability’ were used in particular, whereas the antecedents ‘skills’, ‘sharing’, ‘persistence’, ‘caring’ and ‘discretion’ influenced trustworthiness assessments only after extensive collaboration. The antecedents ‘communality’ and ‘responsibility’ were stressed at both the initial and subsequent collaborative phases. Certain information elements prove to be more informative for professional trustworthiness assessments at the initial phase of collaboration, as they provide multiple cues for the trustworthiness antecedents especially relevant in this phase. |
