|
Article Excerpt
Abstract
This article examines the relationship between interpersonal trust and virtual collaborative relationship (VCR) performance. Findings from a study of 10 operational telemedicine projects in health care delivery systems are presented. The results presented here confirm, extend, and apparently contradict prior studies of interpersonal trust. Four types of interpersonal trust--calculative, competence, relational, and integrated--are identified and operationalized as a single construct. We found support for an association between calculative, competence, and relational interpersonal trust and performance. Our finding of a positive association between integrated interpersonal trust and performance not only yields the strongest support for a relationship between trust and VCR performance but also contradicts prior research. Our findings indicate that the different types of trust are interrelated in that positive assessments of all three types of trust are necessary if VCRs are to have strongly positive performance. The study also established that if any one type of trust is negative, then it is very likely that VCR performance will not be positive. Our findings indicate that integrated types of interpersonal trust are interdependent, and the various patterns of interaction among them are such that they are mutually reinforcing. These interrelationships and interdependencies of the different types of interpersonal trust must be taken into account by researchers as they attempt to understand the impact of trust on virtual collaborative relationship performance.
Keywords: Interpersonal trust, collaboration, virtual teams, telemedicine
**********
Introduction
One should expect trust to be increasingly in demand as a means of enduring the complexity of the future which technology will generate (Luhmann 1979, p. 16).
In a complex world, trust is a necessity. Trust effectively and efficiently reduces complexity by enabling parties with different knowledge bases and experiences to collaborate (Gefen 2000; Luhmann 1979; Lewis and Weigert 1985). Such collaborative relationships can be either virtual or face-to-face. In virtual collaborative relationships, technology is often considered the Achilles' heel; it is likely that trust is the key issue.
This research examines the relationship between interpersonal trust and virtual collaborative relationship (VCR) performance. Findings resulting from data collected in a field study of VCRs in health care delivery, specifically 10 operational telemedicine projects, are presented. VCRs in health care delivery have characteristics that make them exemplary subjects for studying trust in a context where trust is very important, yet they often operate in a context hostile to developing and maintaining trust.
This study makes several contributions to research on interpersonal trust in VCRs. First, the formation of collaborative relationships, be they virtual or face-to-face, does not guarantee they will be effectively utilized. Collaborative relationships are inherently social constructions, and their success or failure may be due to the social context in which they exist, and not the quality of the relationships themselves. This study extends research on the relationship between interpersonal trust and performance. Partial support for a relationship between performance and affect-based interpersonal trust has been found (McAllister 1995), but no relationship between interpersonal trust and performance has also been found (Zaheer et al. 1998). Prior research found that antecedents to trust were a good predictor of trust development in virtual teams (Jarvenpaa et al. 1998), while another study found a relationship between communication processes and swift trust development (Jarvenpaa and Leidner 1999).
Second, this study helps us to understand why technically sufficient information systems may not be adopted. Deploying information technology systems is often predicated on the assumption that the value of an information system's information quality and increased information processing capabilities is sufficient to justify system use. However, this assumption does not always hold, and decision makers often rely on sources they trust the most--regardless of the timeliness, quantity, and quality of information from these sources (Mishra 1996; Staw et al. 1981). There are many reasons why systems are not adopted (Collins and Bicknell 1998). This study attempts to differentiate between complexity reduction efforts that fail due to insufficient information processing capabilities, such as inadequate hardware or software capabilities, and others that fail due to a lack of trust within the relationship.
Third, this study introduces a methodology, facet theory, previously not utilized in information systems research. Facet theory originated in psychology and is a systematic approach to facilitating and integrating research construction, design, and data analysis of complex social systems. It utilizes multidimensional data analysis guided by a theoretical framework (Borg and Shye 1995; Guttman and Greenbaum 1998; Shye 1998). Facet theory has the potential to address many of the concerns and challenges information systems researchers face in performing field research.
Fourth, this research makes a significant contribution to practice by demonstrating that organizations must address interpersonal trust factors if they want to reap the benefits of newer work relationships. It also suggests new considerations in the ongoing concerns about the more extensive adoption of virtual relationships, specifically telemedicine, in the effective and efficient deployment of health care resources.
Interpersonal Trust
Trust in an organizational setting is an effective enabler of complexity reduction--especially when important decisions and new technology are concerned (Gefen 2000; Lewis and Weigert 1985; Ring and Van de Ven 1994). Trust plays a key role as a foundation for effective collaboration (Kramer 1999; Mayer et al. 1995; Rousseau et al. 1998; Whitener 1998) and is the salient factor in determining the effectiveness of many relationships (Gefen 2000; George and Jones 1998; Newell and Swan 2000; Sako 1998; Zand 1972). While trust may reduce transactions costs (Williamson 1981, 1985, 1993), its main impact on collaborative relationship performance is facilitating the learning and innovation (Goshal and Moran 1996; Newall and Swan 2000; Sako 1998) needed to address the ambiguity and unstructured nature of wicked decision problems (Mason and Mitroff 1973). A direct link between trust and collaborative relationship performance exists; once the need for collaboration is established, trust becomes the salient factor in determining performance.
Trust is particularly important in newer organizational forms such as virtual collaborative relationships (McKnight et al. 1998; Meyerson et al. 1996; Newall and Swan 2000; Ring 1996). In the virtual world, trust is a way to "manage people whom you do not see" (Handy 1995, p. 41). Yet trust in virtual teams is difficult to build, and it has been argued that face-to-face contact is irreplaceable for building trust (Nohria and Eccles 1992). The replacement of technology for collocation undermines the emotional relationship aspects of trust. Collocation reinforces social similarity, shared values, and expectations, and increases the immediacy of threats from failing to meet commitments (Jarvenpaa and Leidner 1999; Latane et al. 1995; Sako 1998).
Trust is especially important in health care delivery because health care providers rely on collaboration as a primary means of complexity reduction. Health care delivery is a collaborative activity whose quality, efficiency, and responsiveness is enhanced by the use of interdisciplinary teams (IOM 1996a). Traditionally, health care providers have reduced complexity by collaborating with other health care knowledge workers (IOM 1996a; Silberman 1992). Such complexity results from the co-morbidity of patient conditions that requires providers to simultaneously deal with multiple problems that interact with each other and often preclude assignment of causal relationships with any certainty (IOM 1996b). Patient symptoms and test results are subject to multiple plausible, but conflicting, interpretations (IOM 1996b). Multiple treatment courses with multiple, contradictory, and often uncertain outcomes are available (Silberman 1992), and patient and family members vary in their treatment and quality of life preferences.
While trust is critically important in health care delivery relationships, the health care delivery environment does not facilitate the creation and maintenance of trust (IOM 1996b; OTA 1995;). Collaborative consultations can expose one to malpractice liabilities due to the actions of others. A health care provider engaging in collaborative consultations risks losing patients to the other party, and must trust that the other party will not attempt to steal patients. Health care delivery involves multidisciplinary teams, where team members have to rely on others whose training and perspective are different from their own, and where there are significantly different power and status relationships. Trust in such differentiated teams can be especially difficult to create and maintain (Luhmann 1979). Indeed, it can be argued that lack of trust is the natural state of the health care delivery environment.
The importance of and difficulty in creating and maintaining trust is especially in force in telemedicine. Telemedicine is the process of two or more geographically separated health care providers collaborating via information technology to provide value-added health care delivery (IOM 1996b). The standard measures of health care delivery--and measures of the effectiveness of complexity reduction efforts--are access, cost, and quality (IOM 1990, 1993, 1996a), and it is claimed that telemedicine can increase the access to and quality of health care delivery while simultaneously lowering costs (Allen and Stein 1998; Debakey 1995; Flaherty et al. 1995; GAO 1997; Grigsby 1995; IOM 1996b; ORHP 1997; OTA 1995). Telemedicine can result in significant economies of scale as expertise can be centralized and utilized more frequently (COC 1996; DOC 1997; GAO 1997; IOM 1996b; ORHP 1997).
Telemedicine brings its own additional problems in creating and maintaining trust. Telemedicine projects involve multidisciplinary teams, and virtual processes may differ from what health care providers are used to. Telemedicine blurs individual boundaries and thrusts participants into unfamiliar interactions. Technology changes the mode of presentation on which trust formation is based. Technology may change how and what cues are noticed or ignored relative to face-to-face interactions, and it may force health care providers to interact with team members in a manner different from what they are used to. The installation of telemedicine infrastructure may be interpreted as threatening by remote health care providers, who may see it as the first step toward their replacement by nonlocal providers (GAO 1997). Interpersonal trust may not exist for many telemedicine projects.
While there remain significant variations in conceptualizations of trust by organizational researchers, a general consensus has been reached that trust is a psychological state based on confident expectations and beliefs that another party will act in a certain manner, and that the trusting party must in some way be vulnerable under conditions of risk and interdependency to actions by the other party (Kramer 1999; Rousseau et al. 1998). Our study focuses on the relationship of interpersonal trust with virtual collaborative relationship performance.
Identifying types of interpersonal trust is a contentious and confusing issue. Researchers identify different types of interpersonal trust, use different terminology for similar types of trust, or use similar terms for different types of trust, and subcategorize the same type of trust in different ways. These various subcategories introduce complexity into the study of trust (Bigley and Pearce 1998). For example, cognition trust differs in its definitions and, for some, comprises subcategories of trust that others consider separate types of trust. Some researchers consider calculative trust a core type of interpersonal trust, while others do not consider it a type of trust at all. In this section, we present our model of interpersonal trust and the hypothesized relationships with VCR performance. We have attempted to bring to light the major types of trust in the literature, how definitions of the same type of trust differ, and why we chose the definitions and operationalizations we did.
Figure 1 presents our model of interpersonal trust and the hypothesized relationships with VCR performance. We have identified three types of interpersonal trust--calculative, competence, and relational--and combined them for a fourth type of interpersonal trust--integrated--which includes all three types of trust in our model.
Calculative Trust
Calculative trust is based on conceptualizing trust as a form of economic exchange (Child 1998; Lane 1998; Lewicki and Bunker 1996). Also termed rational trust (Gambetta 1988; Lewicki and Bunker 1996; Mayer et al. 1995; Williamson 1993), calculus-based trust (Rousseau et al. 1998), commitment trust (Newell and Swan 2000), and contractual trust (Sako 1991, 1992), calculative trust is an ongoing, market-oriented, economic calculation where each party assesses the benefits and costs to be derived from creating and sustaining a relationship (Child 1998; Lewicki and Bunker 1996). Calculative trust is a form of contractual agreement where parties can be relied on to deliver according to the details of the contract (Newell and Swan 2000; Sako 1991, 1992). The parties choose whether or not to participate in a trusting relationship based on some form of cost-benefit analysis. Individuals are assumed to be economically rational beings motivated by their desire to maximize expected gains or minimize expected losses in their transactions (Kramer 1999).
[FIGURE 1 OMITTED]
The concept of calculative or rational trust is not without its detractors. Some argue that trust is not the result of conscious calculation (Kramer 1999) nor is it a form of economic exchange (March and Olsen 1989), concluding that calculative trust does not exist. Trust is needed only when conditions of information uncertainty exist; however, calculative trust is effective only when there is little or no need to trust because there are only limited, identifiable conditions of information uncertainty (Child 1998; Lane 1998). Therefore, calculative trust, and the ability to assess the costs and benefits of engaging in a relationship, comes into play only under conditions in which the need for trust is limited.
Despite these objections, we include calculative trust in our model because the condition that the trusting party must be vulnerable to the non-performance of the other party in our definition of trust holds in this case. Trust includes motivational components (Kramer 1999; Shepard and Tuchinsky 1996), and some of these motivational components may be calculative. This leads to our first hypothesis.
H1: There is a positive association between calculative trust and virtual collaborative relationship performance.
Competence Trust
Competence trust is whether the other party is capable of doing what it says it will do (Butler 1991; Butler and Cantrell 1984; Mayer et al. 1995; Mishra 1996; Sako 1991, 1992, 1998). There appears to be a definitional consensus about competence trust in the research community. It is an assessment of the expertise and abilities of the other parties, and is important in a knowledge-based economy because it acts as an indicator of the other party's ability to perform as anticipated (Rousseau et al. 1998). Competence trust is required in complexity reducing collaborative efforts when the skills needed to perform a task are not found within one person (Newall and Swan 2000). A party is more likely to engage in a collaborative relationship if they perceive the individuals in the other party as being capable.
H2: There is a positive association between competence trust and virtual collaborative relationship performance.
Relational Trust
The third type of trust in our model, relational or benevolence trust, is the extent one feels a personal attachment to the other party and wants to do good by the other party, regardless of egocentric profit motives (Jarvenpaa et al. 1998; Mayer et al. 1995). Variations of relational trust include normative trust (Child 1998), goodwill trust (Sako 1991, 1992, 1998), affect-based trust (McAllister 1995), identification trust (Lewicki and Bunker 1996), companion trust (Newall and Swan 2000), and fairness (Zaheer et al. 1998). A motivation to do good by the other party is key to these definitions. These definitions all include one party empathizing with the other party, and specifically excludes the notion of calculative trust.
However, there are significant differences in these definitions as well. Some definitions may or may not include friendship (Lewicki and Bunker 1996; Newell and Swan 2000), affect (Kramer 1999; McAllister 1995; Zaheer et al. 1998), shared identity (Lewicki and Bunker 1996), goodwill (Newell and Swan 2000), common values (Child 1998; Lane 1998), mutual understanding (Lewicki and Bunker 1996) and dependability (Zaheer et al. 1998). Some researchers also include cognitive and motivational underpinnings of relational trust (Kramer 1999). For our purposes, the key to relational trust is that one party empathizes with the other party and wants to do good by them for altruistic reasons. Relational trust is thought to be especially important to the success of collaborative activities (Jarvenpaa and Leidner 1998; Sako 1998). However, prior research (McAllister 1995) has found only partial support for a relationship between performance and affect-based interpersonal trust. This leads to our third hypothesis.
H3: There is a positive association between relational interpersonal trust and virtual collaborative relationship performance.
Integrated Trust
The integrated perspective of interpersonal trust (Lewicki and Bunker 1996; Mayer et al. 1995; Zaheer et al. 1998) combines the different types of trust. The different types of trust are related to each other, even though they may be separable and vary independently of each other (Mayer et al. 1995). Trust can take different forms in different relationships, and different forms of trust may mix together and interact in some situations. Trust may have a bandwidth that can vary in both scope and degree (Rousseau et al. 1998). For example, some relationships may rely more on a combination of calculative and competence trust, while other relationships may be based more on a combination of relational and competence trust. Further, one type of trust may evolve into another, deeper type of trust. Rousseau et al. (1998) speculated that in terms of interpersonal trust, calculative trust was more important in the early stages of a relationship, while relational trust was more influential in the later stages. Other authors have proposed similar ideas, with the initial creation of calculative trust paving the way for the equivalent of relational trust development (Bachman 1998; Child 1998; Lewicki and Bunker 1996; Newall and Swan 2000).
In addition, there is the issue of whether the various types of trust can compensate for each other (Mayer et al. 1995). Can highly positive relational trust compensate for negative calculative trust, and vice versa? Can negative competence trust be offset by highly positive calculative and/or relational trust? While we do not specify a temporal relationship between the different interpersonal types of trust, we argue that the different types of interpersonal trust affect and are affected by each other, and that a combination of the different types of interpersonal trust together impact collaborative relationship performance. However, a prior study (Zaheer et al. 1998) did not find support for a relationship between integrated interpersonal trust and performance. This leads to our final hypothesis.
H4: There is a positive association between integrated trust and virtual collaborative relationship performance.
Types of Trust Excluded from the Model
A number of types of trust identified in the literature were excluded from our model because we felt that they either were not a type of interpersonal trust or, given our definitions, were effectively included in the model. Foremost among these was cognition trust (Child 1998), also called cognition-based trust (McAllister 1995), and knowledge-based trust (Lewicki and Bunker 1996). We chose to exclude cognition trust because the definitions of it were often contradictory or overlapped with the definitions in our model. Some authors included dependability (McAllister 1995), reliability (McAllister 1995), and competence and/or benevolence (Jarvenpaa et al. 1998; Mayer et al. 1995; McKnight et al. 1998) within their definitions of cognitive trust.
Trust as a psychological state implies that all trust is cognitive in that it is based on beliefs about another party. The types of trust in our model (calculative, competence, relational) are all cognitive in that they are judgments or beliefs. The confusing definitions of cognition trust may explain why prior research (McAllister 1995) had difficulty in testing a relationship between cognition trust and performance.
A common aspect of cognition trust is the inclusion of predictability, which is the degree of consistency in intended behavior. It is also perceived as a type of trust in and of itself (Zaheer et al. 1998). We have not included predictability as a type of trust because predictability is not enough to explain trust (Bachman 1998), and to be meaningful, trust must go beyond predictability (Deutsch 1958; Mayer et al. 1995). Instead, we include predictability from both an economic standpoint and a relational standpoint in two of our types of trust. In calculative trust, one party is perceived as predictable if it is in their own self-interest to do so, while in relational trust, one party is seen as predictable because they empathize with the other party.
Some authors include reliability and/or dependability in their definitions of competence trust or cognition trust, where reliability is whether the other party can be relied on to fulfill their obligations (Anderson and Weitz 1989; Zaheer et al. 1998). Reliability is also sometimes perceived as a type of trust in and of itself (Zaheer et al. 1998). Reliability was excluded from our model because professional ethos and fear of malpractice tended to ensure that one party thought the other party reliable prior to agreeing to engage in the telemedicine relationship. It was thus a necessary precondition for the...
|
