Importance of IT and Role Identities in Information Systems Infusion

To develop an understanding of IS infusion behaviors in an organization, it is important to consider an individual’s interactions and perceived relationships with a particular IS in order to find answers to the question, “Who am I, as an individual, through my use of this technology?” (Carter & Grover, 2015).

IT identity is conceptualized as a higher order construct with three dimensions: dependence, emotional energy, and relatedness. Dependence reflects “a person’s sense of reliance upon an IT” (Carter & Grover, 2015, p. 945). Emotional energy refers to “a person’s enduring feelings of emotional energy and enthusiasm in relation to an IT” (Carter & Grover, 2015, p. 945), while relatedness refers to “the blurring of boundaries between the self and an IT, and manifests as feelings of connectedness with the IT” (Carter & Grover, 2015, p. 945). Affective emotional responses to an IT, a strong sense of connection, and successful experiences with an IT reflect stronger identification with a particular IT.

IT identity focuses on self-expansion through exploration and learning until the interaction with IT becomes routinized through the use of a set of IT features. However, as reported in previous studies, some users move from ‘use’ to ‘deep use,’ which requires their commitment and willingness to experiment with IT to extend the frontiers of IT use (Hasan et al., 2016; Kim & Gupta, 2014). These individuals are likely to incorporate the social role of being an IS infuser in an organization into their self-concept. Based on the theoretical argument of role identity, individuals’ IS infusion role identity is likely to form through their self-support, social support, commitment, resource investment, and external and internal gratifications within an organization. Individuals’ self-support regarding IS infusion role identity refers to self-confirming feedback that the infusion behaviors they enact fit their view of infusion role identity. Individuals’ social support regarding IS infusion role identity refers to their belief in colleagues’ perceptions of how they perform as IS infusers and how their colleagues regard their important characteristics as IS infusers. Individuals’ commitment regarding IS infusion role identity refers to how devoted they perceive themselves to be and how committed they are for being the kind of IS infusers that they perceive themselves to be. Individuals’ resource investment regarding IS infusion role identity refers to how much of their working time and energy is devoted to being IS infusers. Individuals’ external gratifications regarding IS infusion role identity refers to their work improvement in performance, speed, productivity, and effectiveness by being IS infusers. Individuals’ internal gratifications regarding IS infusion role identity refers to their positive feeling and enjoyment from being IS infusers.

Next, we present the research models that will be used to evaluate the influence of IT identity and IS infusion role identity on IS infusion behavior and extended use, integrative use, and emergent use.

When individuals have a choice in roles to perform, the meanings attributed to IT use are likely to influence the meanings of their role identities (Burke & Stets, 2009). For example, when a worker identifies her personality with objects associated with her (e.g., technologies), she is more likely to consider those objects to be instrumental in delivering her job tasks with a sense of fulfilment. It is important to recognize that situational context influences the self through shared language and meanings that enable individuals to assume different roles in relation to that specific context (da Costa Netto & Maçada, 2019). Individuals’ IT identity reflects their deep attachment with an object like an IS affects their internalized expectations about competent performance in their work roles that are related to that targeted system usage (McCall & Simmons, 1966; Stryker & Burke, 2000). This is because individuals’ personal characteristics and attachment to an IS is related to the meaning of “who they are” and this meaning influences the meanings of individuals’ role identities in the workplace. In turn, individuals with strong IT identity may choose to claim role identities that align with their IT identity. A study by Mishra et al. (2012) on the care-provision process among physicians found that their reliance and dependence on the implemented IS enabled them to retain and strengthen their autonomy and dominant IS assimilation role identity. Similarly, previous studies have suggested that an individual’s engagement and participation in a digital-based community impacts the formation of the individual’s role identities, as the individual’s identity that is related to a specific IS becomes related to who they are in the roles they perform and the groups they associate with (Carter, 2012; Carter et al., 2017). Therefore, it is expected that individuals’ IS infusion role identity is influenced by their IT identity in the workplace. This leads to the following hypothesis:

When individuals mentally depend on a system to do their job tasks, they are more likely to act proactively; that is, they invest time and effort in engaging in proactive behaviors, such as using the system to its fullest potential. A strong IT identity demonstrates identification – the use of the target IT is integral to a person’s sense of self (who I am) (Carter & Grover, 2015). This indicates that, to investigate IS infusion, it is necessary to assess the influence of individuals’ IT identity in relation to the IS that they use on a regular basis. An individual with strong IT identity seeks more information, increases her knowledge about the IT’s features, and finds out how to apply previously unused features to perform different tasks (Esmaeilzadeh, 2021). For example, Stein and colleagues (2013) found that individuals develop their identity in relation to the systems-in-use, which influence their use of these technologies to accomplish job tasks. In another study, Carter (2012) found that IT identity is motivated by individuals’ desire for self-expansion, and IT identity enactment includes exploring features of a technology and using it in novel ways to improve self-efficacy. Esmaeilzadeh (2021) found that a strong IT identity can inspire users of smart devices to explore previously unused features and embedded resources and capabilities that can be utilized for additional health-related tasks. Through perceiving the benefit and impact of the target system, individuals with a strong personal preference and salient IT identity will have more passion, reliance, and enthusiasm toward using the targeted system and are therefore more likely to extend the usage of the targeted system to reach a better outcome (extended use). Furthermore, they may use the target system to reinforce the linkage among related job tasks to improve their task accomplishment (integrative use), or, when confronted with unexpected issues, individuals with a salient IT identity may explore and find innovative ways of using a target IS (emergent use). On this basis, it is expected that there is a positive relationship between individuals’ IT identity and their use of an IS to its fullest potential within an organization. Thus, we hypothesize:

Individuals are motivated to use an IS if that system can support them to improve their role identity performance (Armitage & Conner, 1999). Drawing on role identity theory, individuals’ IS infusion role identity indicates their internal expectations regarding the use of an IS to its fullest potential, which subsequently influence them to be proactive in using IS features to perform a number of job tasks, enhance linkages among job tasks, and explore new ways of using IS to support job tasks (Burke & Stets, 2009; McCall & Simmons, 1978; Stryker & Burke, 2000). Mishra et al. (2012, p. 740) argue that a “[p]hysician’s role and social identities together constitute their professional identity.” In their study, they focused on the professional role identity of physicians as care providers and the professional social identity of physicians as part of the extended community of care and reported that these identities influence physicians’ IS use behaviors. As IS plays an increasingly important role in people’s work lives, people’s roles become increasingly inseparable from their interactions with IS (Carter & Grover, 2015). In other words, if an individual receives self-confirming feedback on the fit between her enacted behavior and her view of IS infusion role identity (self-support), and also receives support from others (social support), commits to the use of an IS, invests significant effort, time, and energy (resource investment) when using the IS, receives extrinsic rewards such as a bonus or promotion, and receives intrinsic rewards such as satisfaction and enjoyment from using the IS (external and internal gratifications), then it will likely impact on her role identity prominence and influence her IS infusion behavior. When an individual is committed to the use of an IS, she is more able to appreciate the benefits of IS usage and more likely to engage in using more features of the system (Wang & Hsieh, 2006). She is also better able to interconnect multiple job tasks than those who have weak role identity in relation to IS (Wang et al., 2008), and to show more commitment and competency in the role, therefore expending effort in trying out innovative features of IS. We therefore hypothesize:

The context in which information systems in general and enterprise systems in particular are deployed can affect system use (Burton-Jones & Gallivan, 2007) and infusion behaviors (Hsieh et al., 2011). Therefore, we should appropriately contextualize the influence of role identity and IT identity, which were initially developed for a broad set of behaviors (Burke & Stets, 2009; McCall & Simmons, 1978; Stets & Serpe, 2016) and information technologies (Carter & Grover, 2015), before using them to understand IS infusion behaviors in organizations. In addition, we do not know enough about the interplay between IT identity and other contextual characteristics such as individual differences (Carter et al., 2020b). By incorporating context into our theory development, we can generate rich theoretical and practical insights (Hong et al., 2014; Johns, 2017).

We consider the guidelines offered by Hong et al. (2014) for incorporating context into the theorizing of IS phenomena. To identify salient contextual factors, we begin with Burton-Jones and Gallivan's (2007) suggestion that researchers should account for elements of IS use, i.e., user, system, and task when building theories of system use. To further specify pertinent contextual variables, we turn to status characteristics theory (Berger et al., 1972) which helps explain how social structures influence the identity verification process whereby individuals actively seek out opportunities aimed at keeping their perceptions of how others view them congruent with their self-view (Burke & Stets, 2009). Different status characteristics (e.g., age, gender, education) are linked with the verification of identities (Burke et al., 2007; Stets & Harrod, 2004; Stryker & Burke, 2000). This is because one’s position in the social structure becomes important in a situation when it is perceived to be relevant to expectations about performance associated with a particular identity. For example, Stets and Harrod (2004) found that high-status individuals have access to resources which help them in their identity verification process. Burke et al. (2007) found that higher status individuals based on gender (males) aid their verification process of leadership identity because women are typically not viewed as competent or provided equal resources compared to men.

Since we study professionals in the workplace, their IT identity and role identity tend to involve a task orientation that requires competence, knowledge and skills needed to do well in their jobs (Stets & Harrod, 2004). Therefore, we theorize that three personal characteristics (gender, age, education) and three work-related characteristics (position or role in an organization, tenure, and former IT experience with business applications) may play a role in shaping the relationship between identities and IS infusion. Gender carries general expectations for competence and skills. In a workplace context, men, due to their highly task-oriented nature, tend to hold high positions on the status characteristic of gender and will find it easier to verify their IT identity and role identity than females (Burke et al., 2007). Therefore, we expect that the relationship between IT identity and IS infusion and between IS infusion role identity and IS infusion will be especially salient for men. Although age as a status characteristic generally designates competence, increased age has been found to be associated with challenges in processing complex stimuli, difficulty with paying attention to information on the job (Venkatesh et al., 2003) and declines in cognition that become impediment to technology use (Carter et al., 2020a). Thus, we expect that the relationship between IT identity and IS infusion and between IS infusion role identity and IS infusion will be stronger for those who are younger. Education is important with the verification of task-oriented identities such as IT identity and role infusion identity. The status characteristic of education indicates that individuals with high education have more knowledge and skills than those low in education (Stets & Harrod, 2004). Therefore, we expect that the influence of IT identity and IS infusion role identity on IS infusion will be moderated by education, such that the effect will be stronger for those with higher education. Individuals’ positions or roles in an organization such as those in customer-oriented or front-line positions may have more opportunities to develop competence in their application of IT to situations and problems that arise on a daily basis. Thus, we expect that the relationship between IT identity and IS infusion and between IS infusion role identity and IS infusion will be especially salient for employees who hold customer-oriented or front-line positions. Finally, tenure and IT experience with business applications in task-oriented identities such as IT identity and IS infusion role identity signify competence, resources, knowledge, and skills. Therefore, we expect that the relationship between IT identity and IS infusion and between IS infusion role identity and IS infusion will be stronger for those with longer tenure and those with more experience with business applications.

It is important to recognize that there is the potential for a three-way interaction between personal and work-related contextual characteristics and identity. This is because several status characteristics (e.g., education, gender, organizational position) may be salient in a given situation. The theory argues that individuals are likely to combine the implications of each of these salient characteristics and their relevance to the task at hand (Berger et al., 1972; Ridgeway & Smith-Lovin, 1999). However, researchers (Dawson & Richter, 2006; Jaccard & Turrisi, 2003) have suggested that a three-way interaction can either be probed empirically or proposed in a hypothesis with a full explanation of the specification of its manifestation (Dawson, 2014). Since there is no definite theoretical reasoning to help us specify the exact form of interaction, we choose to perform post hoc probing on three-way interactions and report the results, if applicable.

Following the guideline offered by Hong et al. (2014) and theoretical arguments supported by empirical evidence from research at the intersection of identity theory and status characteristics theory, we incorporate three personal and work-related contextual characteristics as moderators of the relationships between IS infusion role identity and IS infusion and between IT identity and IS infusion. This approach allows us to explicate the interplay among identity variables and contextual variables and improve the explanatory power of our research models. Therefore, we hypothesize:

We used measurement items that have been validated in previous studies. A seven-point Likert scale (strongly disagree, disagree, slightly disagree, neutral, slightly agree, agree, and strongly agree) was used to measure all of these key constructs. The measurement items are presented in Appendix Table 7.

In order to fine-tune the survey, the questionnaire was refined in two stages: pre-test and pilot study (Straub et al., 2004). For the pre-test, five domain experts were approached for their comments on the face validity of the measurement items and the clarity of the questions. Several refinements were then made to improve the flow and structure of the questions according to these experts' feedback. The pilot test was conducted at a company that had used a CRM system for more than three years. Findings of the pilot study from 42 respondents indicated that there were no major difficulties in understanding the questionnaire items and instructions. For this study, indicator reliability of 0.70 or higher was accepted. The measurement item analysis resulted in deleting one item related to external gratifications (EXTG5 with indicator reliability of 0.48) to ensure the construct validity. The elimination of this item follows the guideline that an item can be eliminated only when the indicator’s reliability is low and the elimination of that item causes an increase in composite reliability (CR) (Gefen et al., 2000). In this case, the elimination of EXTG5 increased CR from 0.910 to 0.960. We then assessed the reliability of the remaining items: the Cronbach’s alpha coefficients of all constructs were > 0.7, which met the guideline of Gefen and colleagues (2000). These results showed that the survey items were reliable and valid.

In the field survey, we received 510 responses with 413 responses remaining after removing incomplete responses. The data were collected from employees working in the marketing and sales teams of two major telecommunication companies in New Zealand. The CRM system most used by respondents was Salesforce CRM (68.3%). A summary of the respondents’ demographic information is presented in Table 1.

The validity and reliability of the measurement models (Models A and B) were assessed through internal consistency reliability, indicator reliability, convergent validity, and discriminant validity of the measurement items (Chin, 2010). In this study, IT identity is considered as a second-order reflective construct, reflecting three distinct but highly correlated dimensions, while IS infusion role identity is modeled as a second-order formative construct with six reflective first-order factors. Overall, internal consistency reliability, indicator reliability, and convergent validity were tested and reported a satisfactory level (Appendix Table 8).

Discriminant validity was assessed based on the newly introduced criterion for establishing discriminant validity in PLS-SEM, known as Heterotrait-Monotrait (HTMT) criterion (Hair et al., 2019). Typically, for conceptually similar constructs, HTMT values above 0.9 would suggest the lack of discriminant validity between the constructs and with respect to conceptually distinct constructs, HTMT values less than 0.85 indicate discriminant validity (Henseler et al., 2009). All the HTMT values in different models were then less than 0.85, implying the establishment of discriminant validity based on HTMT_0.85 criterion. Cross-loadings of the items in all models and HTMT_0.85 tables are provided in Appendix Tables 9, 10, 11, 12. Correlations from the pooled analysis for both models are reported in Tables 2 and 3.

The item weights were examined for IS infusion role identity, which is a formative construct. Generally, formative constructs have lower absolute values in item loadings compared to reflective constructs (Karimi et al., 2007). This study used the approach suggested by Petter et al. (2007) to assess formative construct validity, which involves testing multicollinearity among the indicators of the formative construct. The formative constructs of IS infusion role identity had weights ranging from 0.11 to 0.45. The VIF ranged from 2.333 to 2.998, which are lower than 3.3, and all were statistically significant with t-statistics from 6.03 to 21.83, p < 0.001 as shown in Appendix Table 13, indicating that multicollinearity is not an issue among formative indicators. In addition, we evaluated the VIF on the interactions between moderators and independent variables. The results showed that all the VIFs ranged from 2.20 to 4.411, which are below the recommended threshold of 5 (Appendix Table 14), indicating that multicollinearity is not a concern in our models.

To reduce the potential for common method bias (CMB), we followed procedural guidelines established in the literature (MacKenzie et al., 2011). Overall, the results supported that CMB was not a significant issue for this study. The implemented procedural and statistical approaches are presented in Appendix Table 15.

To determine the statistical significance of the path coefficients, we ran the bootstrapping method using the number of samples as 2,000 and the number of cases as 300. Most main relationship hypotheses were supported, and the details of the supported hypotheses are presented in Table 4.

IT identity explains 60% of the variance of IS infusion role identity. Taken together, IT identity and IS infusion role identity explain 69% of the variance in IS infusion behavior. Moreover, IT identity and IS infusion role identity explain 63%, 61%, and 57% of the variances in extended use, integrative use, and emergent use respectively. We discuss these findings in detail in the next section.

In addition to evaluating the magnitude of the R² values as a criterion of predictive accuracy, we also applied the predictive sample reuse technique (Q²) to evaluate predictive relevance using a blindfolding procedure. The predictive relevance of IS infusion, extended use, integrative use, and emergent use were obtained using a two-stage approach with the values of 0.577, 0.553, 0.509, and 0.420, respectively. These results show that the models have predictive relevance (Chin & Dibbern, 2010).

Our results show that H1 was supported (path coefficient = 0.779, t-value = 29.703, p-value = 0.000), validating the notion that individuals’ IT identity significantly shapes their IS infusion role identity. The result indicates that individuals’ IT identity (as a set of meanings that individuals attach to the self in relation to a particular IT, and the interactions with that specific IT) affects the extent to which they internalize their self-perceptions about the expectations of using an IS to its fullest potential to perform job tasks (McCall & Simmons, 1966; Stryker & Burke, 2000), thereby supporting their role identity in the IS infusion process.

H2 (path coefficient = 0.409, t-value = 7.601, p-value = 0.000), H2a (path coefficient = 0.470, t-value = 6.845, p-value = 0.000), H2b (path coefficient = 0.321, t-value = 5.977, p-value = 0.000), and H2c (path coefficient = 0.208, t-value = 2.501, p-value = 0.000) were also supported, indicating that individuals’ IT identity shapes their IS infusion behaviors. This is in line with the tenet of identity theories (Burke, 2004; Stryker, 1980) as discussed earlier. In this study, individuals’ IT identity as material identity is guided by their own personal goals, self-interests, and values in relation to IS use rather than expectations or goals of the group or their role in the workplace. The findings of this study suggest that individuals who feel a sense of connectedness, reliance, confidence, and enthusiasm when thinking about themselves in relation to an IT (i.e., strong IT identity) are more motivated to pursue infusion behaviors as well as extended, integrative, or emergent IS use.

When scrutinizing the different aspects of IS infusion, our study’s findings reveal that individuals’ IT identity has a greater impact on extended use than on integrative use and emergent use in the workplace. In other words, individuals who have stronger reliance, dependence, and attachment to the system are more likely to use most features of the system in a deeper manner to do their daily job tasks, as opposed to making innovative and exploratory use of the system.

In addition to the influence of IT identity, we found support for H3 (path coefficient = 0.451, t-value = 8.303, p-value = 0.000), H3a (path coefficient = 0.286, t-value = 4.471, p-value = 0.000), H3b (path coefficient = 0.377, t-value = 6.244, p-value = 0.000), and H3c (path coefficient = 0.478, t-value = 6.013, p-value = 0.000), regarding the influence of IS infusion role identity on IS infusion behaviors. This is in line with findings from Finkelstein and Penner (2004) who examined role identity and organizational citizenship behavior. Our results showed that individuals who personally view the use of an IS to its fullest potential to be an important part of their sense of self are more engaged in IS infusion behavior in the workplace. From a theoretical perspective, individuals who see their IS infusion role identity as being in a higher position in their role identity prominence hierarchy are likely to engage in IS infusion behavior beyond the prescribed or formal guidelines for the use of system features within an organization. When individuals receive strong support from their own self and others on their role identity, as well as commitment, resource investment, and external and internal gratifications in relation to IS use in the workplace, they are likely to be motivated to engage in using more features of an IS or explore the new features of the system in an innovative fashion or reinforce the linkages among multiple job tasks.

The findings show that individuals’ IS infusion role identity has a greater impact on their IS emergent use than on extended and integrative use; that is, these individuals are more engaged in exploring the new features of the system. Thus, individuals with strong IS infusion role identity show more commitment and competency in their role and expend effort in trying out innovative features of an IS. This result is consistent with previous studies that have reported that employees’ IS-related role identities positively influence their IS continuance use behavior in organizations (Farmer et al., 2003; Finkelstein & Penner, 2004).

However, the results of testing hypotheses associated with the influence of contextual characteristics as moderators (gender, age, education, position in organization, tenure, and former IT experience) on the relationships between IT identity and IS infusion, extended, integrative, and emergent use as well as the relationships between IS infusion role identity and IS infusion, extended, integrative, and emergent use, were not all supported. Education is the only personal characteristics that shows significant moderating effects. In particular, H6e is supported (path coefficient = 0.101, t-value = 2.258, p-value = 0.024), suggesting that education moderates the relationship between IS infusion role identity and IS infusion. H6f is supported (path coefficient = 0.112, t-value = 2.589, p-value = 0.010 and H6h is supported (path coefficient = 0.159, t-value = 2.108, p-value = 0.036), indicating that education moderates the relationships between IS infusion role identity and extended use and emergent use respectively. These results suggest that employees with higher education who possess strong IS infusion role identity are endowed with more resources and are more committed to using a CRM system to its fullest potential and more likely to engage in extended and emergent use of these systems. The remaining personal characteristics (i.e., age, gender in H4a-h and H5a-h) had no significant moderator relationships between IT identity, IS infusion role identity and IS infusion, extended use, integrative use, and emergent use behaviors. According to status characteristics theory, age and gender are considered diffuse status characteristics that signify general competencies and may not exert influence in IS infusion behaviors compared to other specific status characteristics (e.g., education, tenure, former IT experience) that are more closely associated with knowledge and skills required to engage in IS infusion (Berger et al., 1972; Stets & Harrod, 2004). The three work-related contextual characteristics have moderating effects on selected relationships between IT identity and IS infusion behaviors and between IS infusion role identity and between IS infusion behaviors. In particular, we found that H7b is supported (path coefficient = 0.193, t-value = 2.836, p-value = 0.005), indicating that an individual’s position in the organization moderates the relationships between IT identity and extended use. H7f is supported (path coefficient = -0.232, t-value = 3.507, p-value = 0.000), and H7h is supported (path coefficient = -0.249, t-value = 2.665, p-value = 0.001), suggesting that organizational position moderates the relationship between IS infusion role identity and extended use and emergent use. These results indicate that employees with more customer-related and front-line team positions (e.g., customer service representative/managers) who embrace strong IT identity and IS infusion role identity are dependent on the CRM systems or are committed to the use of CRM systems for better engagement and greater use of the features of CRM systems. H8d is supported (path coefficient = -0.248, t-value = 2.577, p-value = 0.011). The negative moderating influence of tenure on the relationship between IT identity and emergent use of CRM systems, suggesting that newly joined employees who embrace strong IT identity and IS infusion role identity are more dependent on the CRM systems and are more committed to fully using the systems and trying the new features of these systems. This result contrasts with our hypothesized role of tenure. This may be because those who have longer experience in their positions may develop habits of IT use that lead to fewer attempts at developing deep use of a system (Carter et al., 2020b). Finally, H9f is supported (path coefficient = 0.101, t-value = 2.083, p-value = 0.033), showing a positive moderation effect of former IT experience on the relationship between IS infusion role identity and extended use of CRM systems.

Figure 2 in Appendix shows the interaction diagrams that provide a more nuanced understanding of the interaction of personal and work-related characteristics on the relationships between individual identities and IS infusion.

Due to the non-significant moderating effect of some contextual characteristics on the relationships between IT identity, IS infusion role identity, and IS infusion, extended, integrative, and emergent use behaviors, we further investigated if there are differential impacts of these relationships amongst employees who have different levels of former IT experience and CRM experience. Employees with more IT experience or CRM experience may show a higher level of attachment, dependency, and commitment to engage in IS infusion behaviors compare to employees with less IT and CRM experience.

We conducted a multi-group analysis (MGA) of former IT experience and CRM experience to detect any significant differences in path coefficients of the relationships between IT identity, IS infusion role identity, and IS infusion behaviors. We followed the approaches suggested by Henseler et al. (2009) to assess the variances in population parameters based on the differences in path coefficients and t-values. The parametric test results indicated that there were significant differences in the relationships between IT identity and IS infusion, extended, and integrative use behaviors among employees with more than three years CRM experience and employees with one to three years experience with CRM. The parametric test results also showed that there were significant differences in the relationships between IS infusion role identity and IS infusion, and extended use behaviors among employees with more than three years CRM experience and employees with one to three years experience with CRM. Moreover, the parametric test results indicated that there were significant differences in the relationships between IT identity, IS infusion role identity and IS infusion and extended use behaviors among employees with more former IT experience than employees with less former IT experience. The MGA results are presented in Appendix Table 16.

Moreover, we tested the mediator role of IS infusion role identity between IT identity and IS infusion, extended, integrative, and emergent use behaviors. From the analysis, IS infusion role identity was found to influence IS infusion, extended use, integrative use, and emergent use positively, and also IS infusion role identity was positively influenced by IT identity. In particular, based on Baron and Kenny’s (1986) guidelines, IS infusion role identity partially mediated the relationship between IT identity and extended use, integrative use, and emergent use behaviors. Subsequently, Sobel’s test was used to assess the significance of the mediating relationship, which showed that IS infusion role identity was a significant mediator. The results are presented in Appendix Table 17.

This study contributes to both research and practice. Drawing on identity theories, this research offers insights into the role and importance of IT identity and role identity in IS infusion behaviors in organizations. In particular, the findings reveal that individuals’ IT identity as their material identity strongly influences their IS infusion role identity within an IS use context. In this case, individuals who have strong IT identity are likely to choose roles consistent with these meanings to give them more opportunities to interact with the IS that they view as integral to their self-concept.

The study further extends our understanding that individuals’ IT and role identities have different degrees of influence on the three ways that demonstrate IS infusion behaviors. In particular, IT identity has a stronger influence on extended use and integrative use while IS infusion role identity exerts a stronger influence on emergent use. This means that both IT identity and IS infusion role identity play a complementary role in promoting the overall IS infusion behaviors that involve using a number of IS features, new feature exploration, and novel use of IS features. This study therefore finds that individuals’ identities are important drivers of their IS infusion behavior, be it their extended, integrative, or emergent IS use.

In addition, this study also points out that employees’ personal and work-related characteristics play a role in IS infusion behaviors, thus helping to enrich our understanding of employees’ infusion behaviors in an organization. In particular, some of the salient work-related characteristics that strengthen the influence of IT identity and IS infusion role identity on IS infusion behaviors are IT experience in general and CRM experience in particular, and whether employees are in customer-related or front-line roles. Education as a personal characteristic also shapes the influence of IS infusion role identity on IS infusion behaviors.

Methodologically, this study operationalized IT identity and IS infusion role identity as second-order constructs through their sub-dimensions within an organizational context. IT identity was operationalized through three reflective dimensions: dependence, emotional energy, and relatedness. IS infusion role identity was operationalized through six formative determinants: self-support, social support, commitment, resource investment, and external and internal gratifications. Most previous studies have examined the role identity construct as a single-dimensional construct through only a few measurement items (Farmer & Van Dyne, 2010; Farmer et al., 2003). Although both single-dimensional and multi-dimensional constructs are appropriate to measure identity, our study shows that the operationalization of IT identity and IS infusion role identity as multidimensional constructs offers rich understanding of their influence on IS infusion behaviors in an organization. Future research may want to adapt these measurement items to suit their study contexts of IS use in organizations.

Table 5 presents a summary of the unique findings of this study in relation to the current body of knowledge in the literature.

In terms of practice, given the impact of IT identity and IS infusion role identity in IS infusion, extended use, integrative use, and emergent use behaviors, organizations can help to promote individuals’ IT identity and IS infusion role identity by providing them with opportunities to use IT to coordinate among job tasks, and by implementing mechanisms to support and reward individuals’ attempts to explore new features of the system. One possibility to strengthen individuals’ IT identity and motivate them to depend on the target system to do their daily job tasks is to design customized training interventions. These interventions should focus on developing individuals’ understanding of organizational improvement goals and IS investment through their effective and full use of the system.

The analysis revealed that social support and external gratifications have a strong influence on individuals’ role identity in relation to IS use. Therefore, managers should outline their expectation and provide appropriate resources to encourage IS infusion behavior. In addition, organizations should develop strategies and conditions to reassure individuals that their IS infusion behavior is supported and rewarded and thus reinforce individuals’ IS infusion role identity. An organization that values IS infusion behavior should provide opportunities and a working atmosphere in which peers are encouraged to reflect on each other’s IS usage behaviors, talk about their expectations, and discuss their concerns about the system and its impacts on work. Managers should identify those individuals with strong IS infusion role identity and place them in positions in which they can act as super users in order to provide these individuals with enhanced status and opportunity to influence other users.

While the findings of this study provide insightful contributions for research and practice, certain limitations should be acknowledged. The methodology adopted, i.e., a cross-sectional survey design and a quantitative approach, may limit the interpretation of the results. With the intertwined nature of the relationship between identity and behavior, future studies should (i) investigate how individuals’ IT identity and IS infusion role identity change over time as they continue to interact with IS; and (ii) establish the cause and effect, while also examining the possible reciprocal causation in the research models longitudinally. Future studies may apply qualitative research methods to provide explanatory insights and underlying meanings regarding how individuals’ IT identity and IS infusion role identity shape their IS infusion behaviors.

It is also plausible that the results may have been impacted by sample selection bias. Future studies may address this point by testing the theoretical models in other systems and contexts. Since individuals can hold many identities simultaneously, future research should further investigate the relationships between IT identity, IS infusion role identity, and other workplace identities (e.g., social identity, professional identity) and IS infusion behaviors.