An Investigation on the Effects of Perceived Transformational Leadership and Perceived Computer Self-Efficacy on Adaptive System Use

An Investigation on the Effects of Perceived Transformational

Leadership and Perceived Computer Self-Efficacy on Adaptive

System Use

MSc Business Administration - Change Management

University of Groningen

Faculty of Economics and Business (FEB) Department of Innovation Management and Strategy

Supervisor:

dr. Ileana Maris-de Bresser Co-assessor:

dr. Hille Bruns

January 2017

ABSTRACT

This thesis looked to explore, through the lens of social-cognitive theory, how perceived transformational leadership affects the behaviours and triggers of Adaptive System Use. In addition, it explores how an individual’s computer self-efficacy plays a role in this interplay between the aforementioned concepts. The study discovered two types of triggers of Adaptive System Use through transformational leadership - primary triggers and secondary triggers. Primary triggers directly cause Adaptive System Use behaviours as long as the individual is willing and able to act upon the trigger, while secondary triggers serve as a kind of groundwork for the primary triggers, unable to directly trigger Adaptive System Use by themselves but able to facilitate and activate primary triggers. Secondly, the study discovered the mediating and triggering effects of computer self-efficacy upon these behaviours, and the effects of differing levels of self-efficacy upon the strength of the triggers. Individuals with higher levels of computer self-efficacy were able to overcome and use limitations in the software as a springboard to perform Adaptive System Use behaviours, while users with lower computer self-efficacy were prevented from acting upon computer self-efficacy related triggers by their self-efficacy. The study resulted in a conceptual model of the effects of transformational leadership and computer self-efficacy on Adaptive System Use.

Keywords: Social-cognitive theory, adaptive system use, computer self-efficacy, transformational leadership, post-adoptive system use.

ACKNOWLEDGEMENTS

INTRODUCTION ... 2

THEORETICAL FRAMEWORK ... 7

Adaptive System Use... 7

Leadership in adaptive system use: Transformational and Transactional ... 10

The role of self-regulation in the creation of individuals’ Self-Efficacy ... 12

Computer Self-Efficacy ... 12

Towards a theoretical understanding of the effects of leadership and self-efficacy on Adaptive System Use: A Social-Cognitive Perspective ... 14

METHODOLOGY ... 16

Research Approaches ... 16

Research setting: The case study in this research ... 16

Data Collection ... 18

Characteristics of Interviewees ... 19

Data Analysis ... 19

Quality Criteria for Research ... 20

FINDINGS ... 22

Adaptive System Use Behaviours from Sun (2012) ... 22

New Adaptive System Use Behaviours ... 23

Triggers of Adaptive System Use ... 24

Computer Self-Efficacy and Triggers of Adaptive System Use ... 30

Computer Self-Efficacy as a Mediator between Transformational Leadership and Adaptive System Use Behaviours ... 32

DISCUSSION ... 34

Answering the Research Questions: What effects do Perceived Transformational Leadership and Computer Self-Efficacy have on an individuals’ Adaptive System Use? ... 34

Secondary Triggers of Adaptive System Use Behaviours ... 38

Primary Triggers of Adaptive System Use from Transformational Leadership ... 39

Computer Self-Efficacy as a Trigger and a Barrier to Adaptive System Use Behaviours ...41

CONCLUSION ... 43

Practical Implications ... 43

Limitations and Directions for Future Research ... 44

Conclusion ... 44

REFERENCES ... i

APPENDICES ... ix

Appendix A: Organizational Chart of IT Department ... ix

Appendix B: Interview Protocol – Managers/Team Leaders/Supervisor ... x

Appendix C: Interview Protocol – Employees/Colleagues ... xiii

Appendix D: Codebook ... xvi

Appendix E: Transcript of Interviews ... xxvi

INTRODUCTION

In recent years, increased reliance on IT has permeated all aspects of society. Increased information flows have improved our responses to natural disasters (Day, Junglas, & Silva, 2009) and shaped our education systems through platforms such as virtual learning (Park, 2009). This same trend has occurred in the private sector (Heroux & Fortin, 2014; Hitt & Brynjolfsson, 1996) where information systems are no longer tools to be used to find an edge over the competition (Kearns & Lederer, 2004; Wilkin & Chenhall, 2010), but are now critical to the day-to-day operation and success of organizations (Alavi & Joachimsthaler, 1992).

The criticality of the hard and soft aspects of IT systems has become apparent. Research into human-computer interaction has become an area of substantial interest (Meho & Rogers, 2008) within healthcare (Sarcevic, 2007), education (Crook, 1998; Couse & Chen, 2010) and business (Townsend, DeMarie & Hendrickson, 1998; Wang, Bakker, Wagner & Wakefield, 2007).

One model stemming from this field of human-computer interaction, developed with the intention to map how users change how they use technology is Adaptive System Use (Sun, 2012). Sun (2012) defines Adaptive System Use as:

“...a collection of specific behaviours that one performs in order to revise his/her use of information system features” (Sun, 2012, p.457).

Figure 1: Conceptual Model of Adaptive System Use by Sun (2012)

The key argument from Sun’s study was that the nature of revisions of system use from users in the post-adoptive stage of technology implementation was understudied. More so Sun argued that the ‘how’ and ‘why’ questions relating to how these revisions are made remained unanswered (Sun, 2012). Conclusively Sun found that users revise their system use in response to triggers, with the moderating effect of contextual factors. Sun (2012) also found individuals perform different behaviours in response to different triggers.

Relating to change management concepts on how individuals respond to stimulation from leaders (Balogun & Johnson, 2005; Bouckenooghe, 2010), examining possible effects of leadership on Adaptive System Use behaviours is an interesting concept that this research hopes to unearth. The first objective of this study therefore is to explore the effect of perceived transformational leadership on individuals Adaptive System Use.

Although Sun found facilitating conditions and deliberate initiatives had no impact on Adaptive System Use behaviours directly, others oppose this view, suggesting leadership, in particular perceived transformational leadership, (a potential trigger not examined by Sun) can positively influence subordinates behaviours and responses in a general sense.

Perceived transformational leadership has become widely researched, extensive attention has been paid to the role of perceived transformational leadership in achieving organizational effectiveness (Judge & Piccolo, 2004; Yukl, 2012), improving innovation and performance (Sun, Xu, & Shang, 2014) and employee commitment (Podsakoff, MacKenzie, & Bommer, 1996).

Transformational leadership maintains users may be inspired to seek new challenges and perform at a higher level of self-efficiency (Waldman, Ramirez, House, & Puraman, 2001) and better deal with environmental uncertainty (Bass & Avolio, 1989) due to support and resources provided by transformational leaders. According to Sarin and McDermott (2003) and Ford and Seers (2006) transformational leadership can involve enlightening activities that encourage use of innovative technical methods and implementing feedback mechanisms to learn from one’s errors, while Zhou (2013) proposed supportive management could enhance the ability of individuals to work creatively.

Although these aspects of perceived transformational leadership theoretically fit within Sun’s (2012) definition of facilitating conditions, it was discovered that facilitating conditions did not moderate the relationship between the triggers and Adaptive System Use behaviours (Sun, 2012).

Reasoning why this is, Sun’s (2012) measurement of facilitating conditions was admittedly not robust enough to cover the entire construct. The study’s use of Microsoft Office, a mature technology that is considered easy-to-use, to test their hypothesis may also have affected the results. Additionally, the effects of leadership on subordinates’ responses were not considered.

leadership can affect responses and behaviour in subordinates and is relevant to examining the specific effects of perceived transformational leadership on Adaptive System Use. Cho et al., (2011), delved into transformational leadership within the realm of IT, examining effects of perceived transformational leadership upon IS success through two psychological mechanisms – perceived organizational support and systems self-efficacy. Perceived Organisational Support refers to an individual’s perception of how the organisation values their contributions and cares about their well-being (Rhoades & Eisenberger, 2002). Systems self-efficacy relates to one’s belief in their capabilities to utilize the information system (Cho et al., 2011).

However, research in this area - effects of self-efficacy in shaping users’ responses in an information system context - is limited, while also being subject to competing terminology and concepts (Li & Hsieh, 2007). Sun (2012) has also acknowledged the importance of self-efficacy in Adaptive System Use is understudied and suggested that future research should take it into account (Sun, 2012, p.472). This, in conjunction with work of Cho et al. (2011) raised the question of how users’ self-efficacy shapes how or what features of the system they use. The concept of Self-Efficacy is defined as:

“...people's judgments of their capabilities to organize and execute courses of action required to attain designated types of performances” (Bandura, 1986, p.391).

Most research regarding self-efficacy is concentrated within the sociological or psychological fields (Bandura, 1986). Although research exists on self-efficacy within an IT context, it centres on effects of self-efficacy to perform tasks in a non-business context (Igbaria & Livari, 1995), personality traits as antecedents to computer self-efficacy (Thatcher & Perrewé, 2002) and developing computer self-efficacy measures (Compeau & Higgins, 1995; Marakas et al., 1998). Research on its relationship with transformational leadership and Adaptive System Use is limited. For studying the role of computer self-efficacy in this research, Compeau and Higgins’ (1995) adjusted definition of general computer self-efficacy – the “belief in one’s capability to use a computer” was utilized.

“As for internal contexts, more systematic studies on other personal factors such as Computer Self-Efficacy are necessary for a better understanding of the internal context of ASU behaviours”(Sun, 2012, p.472).

The second research objective of this study therefore is to examine the effect of perceived computer self-efficacy on Adaptive System Use behaviours, filling the gap proposed by Sun (2012).

In light of the literature gaps identified and the research objectives outlined above, this research aims to explore the effects perceived transformational leadership may have on triggering certain Adaptive System Use behaviours, as well as whether perceived computer self-efficacy has any effect in triggering Adaptive System Use behaviours, and what that effect might be. The research questions that guide this study are:

RQ1: What effect does perceived transformational leadership have on an employee’s Adaptive System Use?

RQ2: How does computer self-efficacy, in relation with transformational leadership, affect Adaptive System Use behaviours?

THEORETICAL FRAMEWORK Adaptive System Use

Adaptive System Use as conceptualized by Sun (2012) is an attempt to map how and why users revise their use of technologies in the post-adoptive phase. It utilized the research of Louis and Sutton (1991) on switching cognitive gears from automatic thinking, to active thinking, following the suggestion of Venkatesh and Davis (1996) that effective IT use is reliant upon the users’ past experiences with IT. At the heart of Adaptive System Use lies the notion of Features in Use (Sun, 2012). ‘Features-in-Use’ is a new construct defined by Sun (2012) within the scope of Adaptive System Use described as:

“…the basket of system features that are ready to be used by a particular user to accomplish tasks” (Sun, 2012, p.455).

It is derived from the idea that different people use different features while using the same system - the concept intending to define a user’s understanding of the system they use. Yamauchi and Swanson (2010) similarly proposed that individuals confront new systems as a learning process where they experiment through trial-and-error. Sun’s ‘basket of features’ that a user has available to them has a clear intersection with Yamauchi and Swanson’s (2010) ‘familiarity pockets’ when examining how users adapted and expanded their capability to utilize information systems.

According to Sun (2012), it is Features-in-Use that mediate an individual’s interaction with an information system, and defines their understanding of the system. Under the umbrella of Features in Use lie four behaviours, divided into two further categories; Revising the Content of Features in Use, and Revising the Spirit of Features in Use.

Construct Definition

Revising the Content of Features in Use: Revisions regarding “what” features are integrated

in a user’s Features in Use

Trying New Features  Add new features to one’s Features in Use and thus

expanding the scope of Features in Use

Feature Substituting  Replacing Features with other features with similar functions.

Revising the Spirit of Features in Use: Revisions regarding “how” a user makes revisions use

of their Features in Use.

Feature Combining  Using features in one’s Features in Use together for the

first time.

Feature Repurposing  _{Using features in one’s Features in Use in a new way.} Table 1: Dimensions and Sub-Dimensions of Adaptive System Use (Sun, 2012)

Engeström, Miettinen, and Punamäki (1998) suggest that innovativeness necessitates deviations from normal use of information system features. Sun (2012) notes, however, that Adaptive System Use is a risky endeavour as owing to the innovative nature of Adaptive System Use there is no real way to predict whether certain revisions by the user may improve the effectiveness or efficiency of their system use. Furthermore, revisions could also be damaging and may cause further problems compared to the previous way the user utilized the system. Jasperson et al. (2005) highlight that active thinking is an energy and time-consuming process and it is possible both the user and organization could incur opportunity or even financial costs from failed revisions. Sun (2012) also acknowledges Adaptive System Use behaviours and revisions may upset or create conflict between users, following Janssen (2003) who maintains that challenging of established formal and informal practices at the workplace could affect the expectations individuals have of one another, risking conflict.

Looking at behavioural patterns, Sun (2012) identified that there are certain triggers that lead to Adaptive System Use behaviours. Deliberate Initiatives, Discrepancies, and Novel Situations were recognised as three possible triggers.

Burton-Jones and Grange (2013) provided a theoretical background in how to move from merely using a system, to adapting the way they utilize the system to more effectively attain the desired goals for the use of that system. Both of these papers however do not deal with the post-adoption phase of IT implementation.

Returning to the triggers defined by Sun (2012), Novel Situations describe incidences where the user encounters a situation where they are required to perform a task that is unfamiliar or completely new to them (Sun, 2012). Within this sub-construct, there are three subcategories, New Tasks, Others Use, and Changes in System Environment. New Tasks deal with situations where the user is required to perform an unfamiliar task. Sun (2012) follows the theories of Ahuja and Thatcher (2005) that the potential for task overload can lead to users re-evaluating their system use and try more innovative (but also more risky) adaptations to cope with changes. Observing other users’ system use may also trigger Adaptive System Use. Boudreau and Robey (2005) posit that observation of other’s system use, or observing ‘power users’ utilization of the system can also trigger Adaptive System Use behaviours. Changes in System environment (Shaw, 2001; Benamati, Lederer & Singh, 1997) may also cause Adaptive System Use as users struggle to cope with the changes imposed upon them.

The second trigger, Discrepancies, refers to situations where actual outcomes of system use differ compared to expected outcomes (Sun, 2012). Discrepancies occur when users’ experience and capability with the system cannot be assimilated into their existing cognitive schema (Wong & Weiner, 1981). Discrepancies, as with Novel Situations, may trigger users to attempt innovative adaptations to solve the discrepancy. Others have researched similar concepts, Leonard-Barton (1988) investigated the misalignments between information systems and the local environment in which it is used, and Beaudry and Pinsonneault (2005) researched how individuals evaluate and make use of stress and coping mechanisms to make sense of such changes.

The final trigger, Deliberate Initiatives, is the suggestion that Adaptive System Use behaviours occur when one is requested to use the system differently (Sun, 2012). Schön (1983) suggested people switch from automatic to active thinking when challenged with demands. Balogun and Johnson (2005) suggest schema and sensemaking ability are influenced by social interactions with colleagues and leaders that challenge users existing schema.

found to be a direct antecedent of Adaptive System Use, however they were found to be antecedents of discrepancies, supporting the hypothesis that Deliberate Initiatives have an indirect effect on Adaptive System Use. While Sun (2012), through the concepts of Deliberate Initiatives and Facilitating Conditions, hinted at the possible role of leadership in influencing Adaptive System Use, he, did not explore specifically the possible effects of transformational leadership on Adaptive System Use behaviour, the core issue this study examines.

Leadership in adaptive system use: Transformational and Transactional

Transformational leadership, first conceived by Burns (1978) and expanded by Bass (1985) is the concept that the leader can influence their followers beyond their own self-interests and more in line with those of the organization. This is achieved either through charismatic means that spiritually appeals to followers, or by intellectually stimulating them (Bass, 1999). Felfe, Tartler, and Liepmann (2003) proposed that among the central tenets of transformational leadership are visioning, value-based attractions, role-modelling, trust, support of personal growth and consideration of followers needs. In contrast, transactional leadership emphasises the “give and take” exchange between leader and follower to cater to both parties self-interest (Sun et al., 2014; Bass 1985).

Transactional Leadership

job satisfaction and organizational commitment, additionally being indirectly associated with positive job performance outcomes in financial auditing organizations, which make substantial use of information systems. As the current literature indicates that in IT contexts transformational leadership is more effective, this type of leadership will be examined in this study as well.

Transformational Leadership

Bass (1999) suggests transformational leadership has four distinct dimensions. Idealized influence and inspirational leadership are part of the visioning process and are exhibited when leaders show examples to be followed, similar to Felfe et al.’s (2003) idea of visioning and role-modelling being central tenets in transformational leadership. Bass (1999), Sun et al. (2014) and Cho et al. (2011) propose that subordinates thrive when these two dimensions are displayed as they feel a will to identify with leaders’ values. Intellectual stimulation refers to a situation when the subordinate is aided by the leader to become more creative and confident (Bass 1999). Related to this is the fourth dimensions, individualized consideration. Individualized consideration involves being attentive to the developmental needs of subordinates and aiding the process of development through individualized coaching and support (Bass, 1999).

By being a role-model and building rapport with followers, transformational leaders are able to encourage followers to think creatively, independently and develop new ideas and thought patterns (Bass & Avolio, 1990). Others maintain that the followers themselves must be willing and ready to accept the transformational leader and work with them for it to be effective (Ambrose & Kulik, 1999).

The dimensions of transformational leadership are robust and sufficiently researched within the field of management and business studies, sociology and psychology (Gundersen, Hellesoy & Raeder, 2012; Choi, 2006). Regarding transformational leadership in IT however, Li & Hsieh, (2007) note that there has been a lack of studies on transformational leadership in this area. Before delving into this, we must ask why do we seek to research transformational leadership in an IT context rather than transactional leadership?

empowerment; and found evidence to support this hypothesis. The logical next step therefore, is to examine whether the same connections hold true with respect to with transformational leadership, and its effect on Adaptive System Use.

The role of self-regulation in the creation of individuals’ Self-Efficacy

According to Bandura (1986) self-regulation of motivation and performance are governed by a set of interrelated self-regulatory mechanisms.

The term ‘self-regulation’ is a catch-all term used to describe these sets of mechanisms (Bandura, 1991). Self-Regulatory mechanisms consist of a process beginning with self-observation - how people observe their own performances and the effects these performances produce (Bandura, 1991). However, self-observation is not simply a reflection of one’s own performances and functioning. Individuals’ pre-existing cognitive structures and beliefs influence these self-observations (Bandura, 1991). One’s mood can also affect the clarity and quality of these self-observations (Kuiper, MacDonald, & Derry, 1983). The importance of self-observation is that it provides the foundation for goal-setting and evaluation of one’s current position vis-à-vis their goals (Bandura, 1991).

Following self-observations of one’s own behaviour, these observations are compared to the individuals pre-developed personal standards, which are formed by how significant persons in the individual’s life have reacted to their behaviour (Bandura, 1991). In turn, these standards provide the basis for self-reaction, the final stage of the self-regulation process. Self-Reactions regulate whatever course of action the individual may take (Bandura, 1991). Furthermore, self-reactions are affected by external sources, such as potential incentives or rewards, or potential consequences. Notably, individuals who engage in self-rewarding or incentivising are generally more successful in their self-regulation (Zimmerman, 1989; Perri & Richards, 1977).

The relation of the self-regulation mechanisms to this study is that it is the underlying principle that creates an individual’s self-efficacy (Bandura, 1991). Self-Efficacy is key in determining an individual’s choices, their given effort, their desire to succeed, and the levels of stress they can cope with (Bandura, 1991). In an IT context, self-efficacy therefore describes one’s self-belief in their efficacy at IT use.

Computer Self-Efficacy

& Higgins, 1995; Murphy, Coover, & Owen, 1989; Agarwal, Sambamurthy, & Stair, 2000; Marakas et al., 1998; Cassidy & Eachus, 2002).

Murphy et al., (1989) made an early attempt to relate the work of Bandura to IT, developing a scale to measure what they term “perception of one’s capability regarding specific computer-related knowledge and skills” (Murphy et al., 1989, p.893). Although this work was aimed toward the design of a measurable scale for computer self-efficacy, the concept itself was increasingly believed to be important toward our understanding of human behaviour towards technology (Agarwal et al., 2000). Compeau and Higgins (1995) determined that computer self-efficacy was a significant factor in influencing individual’s expectations about the potential outcomes of using computers, along with their emotional reactions to technology. They discovered an individual’s computer self-efficacy was positively influenced by the support and encouragement of other workers within their team or group. This followed the original concept of self-efficacy by Bandura (1986), while also validating that the concept of self-efficacy was applicable to an IT context. Succeeding this, studies, such as Cho et al., (2011) and Sun (2012) have further confirmed the applicability of self-efficacy to an IT context.

Compeau and Higgins (1995) theorized that an individual’s computer self-efficacy moderated the relationship between organizational influences (e.g. pressure from management) and an individual’s own decision of whether to use a given technology. Marakas, Yi and Johnson (1998) further enhanced the concept of computer self-efficacy by proposing that the term ‘computer self-efficacy’ was too general and unsuitable for use, that one catch-all concept could not describe one’s efficacy at using all the functions that a technology has to offer. They further suggested two subcategories within computer self-efficacy: task-specific computer self-efficacy and general computer self-efficacy. Task-specific computer self-efficacy refers to:

“...one’s perception of efficacy in performing specific computer-related tasks within the domain of general computing” (Marakas et al., 1998, p.128).

Task-specific computer self-efficacy applies to use of applications or programs. The advantage of this is that, unlike Compeau and Higgins’ (1995) measure, it allows a more specific measurement of an individual’s ability to use a single function of a technology without harming the quality of the assessment by assessing the individual on a range of functions that may not be relevant to them (Marakas et al., 1998).

“...efficacy across a range of multiple computer application domains” (Marakas, et al., 1998, p. 129).

General computer self-efficacy is more in line with what is traditionally known as computer self-efficacy as defined by Compeau and Higgins (1995). It is this definition, general computer self-efficacy that will be utilized as the definition of “computer self-efficacy” for the purposes of this study.

Towards a theoretical understanding of the effects of leadership and self-efficacy on Adaptive System Use: A Social-Cognitive Perspective

Cho et al. (2011) examined the role of perceived transformational leadership on users’ information systems success based upon two mechanisms – perceived organisational support and systems self-efficacy, and discovered transformational leadership positively associated with information system success, with the relationship mediated by users’ self-efficacy. Cho et al. (2011) also found that the amount of perceived organisational support also was an important factor in information systems success. In their research however, although Cho et al., (2011) made use of elements of Bandura’s (1977) work on self-efficacy, they lack an overarching theory of which to bring these concepts (perceived transformational leadership and computer self-efficacy) together.

For this, we examine the work of Bandura (1977, 1982, 1984, 1986), on social-cognitive theory. This theory, we believe, plays a crucial role in bringing together the concepts of perceived transformational leadership and perceived computer self-efficacy in investigating their effect on Adaptive System Use behaviours.

The importance of this theory in terms of its relevance to this research lies not only in its connections to self-efficacy, but also in its suggestion that environmental factors affect individual’s self-efficacy. Indeed, social-cognitive theory poses that human behaviour is a function of: (1) cognition, (2) personal factors, such as past experiences and (3) environmental influences, such as managerial choices and perceived organisational support, interacting and influencing each other reciprocally (Wood & Bandura, 1989). These factors may not all be of equal strength, and the effects can be seen throughout an individual’s life (Wood & Bandura, 1989). Better explaining our use for Bandura’s theory as a lens for investigating the effects of environmental influences on behaviour is this quote from Bandura, (1997):

human agency operates generatively and proactively rather than just reactively”(Bandura, 1997, p.6).

Taking Adaptive System Use behaviour as a type of human behaviour, social-cognitive theory implies that to be fully understood, the effects of cognitive and personal factors need be known, but also the effects of environmental influences in generating such behaviour (Bandura, 1997). In this study, the environmental influence studied is perceived transformational leadership, and our interest is in establishing its effects on Adaptive System Use behaviour. Personal factors are examined in this study in terms of perceived computer self-efficacy. This lens should provide insights into the effects of computer self-efficacy on Adaptive System Use, filling a research gap mentioned by Sun (2012).

Making use of Bandura’s (1977, 1982, 1984 ,1986, 1997) social-cognitive theory therefore, we bring the concepts of perceived transformational leadership and perceived computer-self-efficacy together to create a lens through which to examine how they affect Adaptive System Use.

Walumbwa et al. (2008) suggested that transformational leadership could positively affect individuals’ self-efficacy through role-modelling and persuasion. Cho et al. (2011) tested this hypothesis and indeed found transformational leadership had a positive impact upon computer self-efficacy. This was consistent with social-cognitive theory, which posits that self-efficacy is subject to the effects of environmental influences (Bandura, 1989, 1997), in this case transformational leadership. Expanding this to Adaptive System Use, social-cognitive theory therefore offers an ideal lens through which to view the interplay of these concepts and their effects on Adaptive System Use behaviours.

METHODOLOGY Research Approaches

The aim of this study is theory development, as we intend to explore connections between perceived transformational leadership, perceived computer self-efficacy and Adaptive System Use. Therefore a qualitative approach, following the suggestions of Eisenhardt (1989), based upon developing theory from case studies was utilized.

The reasoning for choosing this approach has several explanations. Firstly, case studies fit well in the context of research questions that attempt to ask questions or investigate present circumstances (Yin, 2003). A case study may be supplemented by a literature review to provide a foundation for the research, allowing researchers to make comparisons with the results found from the case studies (Yin, 2003). The literature review in the previous chapter serves this purpose. Feagin, Orum and Sjöberg (1991) suggest investigating people in their natural environment, as case studies allow, unearths empirical data that may not appear through other methods of research. Finally, previous studies on Adaptive System Use have used a quantitative approach. Given the lack of exploratory work on Adaptive System Use in general and more specifically on examining the linkages between Adaptive System Use and perceived transformational leadership and computer self-efficacy, a case study approach was deemed a suitable approach to follow.

  However, case-studies also have limitations. Often case study researchers often end up underestimating the amount of data they expect to collect and ‘drown’ in the data (Hodkinson & Hodkinson, 2001). Such issues are prevented by having a suitable, narrowly focused research question prior to data collection (Eisenhardt, 1989).

Research setting: The case study in this research

An embedded single-case study was conducted at the Centre for Information Technology at the University of Groningen, The Netherlands. Specifically, the research was conducted within the Educational and Innovation Support department, a sub-department within this department. Appendix A shows the structure of the department.

such as Slack and ‘fogbugz’. Table 2 below shows a breakdown of these specialist software used by helpdesk support.

Software Name:

Functionality: Use by Support:

Slack Professional chat and messaging application

Used for communication between employees in the department.

Fogbugz Web-based project

management software, with built-in features for bug tracking, wikis, CRM, and scheduling.

Used as a ticketing system to organise queries and emails from staff/students. Also used to provide a wiki of information on requests, bugs and other issues with the software, as well as a ‘how to’ guide for helpdesk support.

Nestor (a.k.a Blackboard)

Virtual Learning Environment and Course Management platform.

Support utilize Nestor in both a staging and real environment, the staging environment is used to test features and tools without damaging the real environment.

TestRail Web-based test management software

Used to track and manage software testing efforts, quality assessments, etc.

and management. The sampling method used therefore was one of theoretical sampling, in line with the research approach, and chosen for the purpose of extending emergent theory (Eisenhardt, 1989).

Data Collection

Following the suggestions of Eisenhardt (1989) and Sun (2012), semi-structured interviews were used to gather data. The nature of semi-structured interviews allows the flexibility to probe and ask further concept-related questions that structured interviews lack (Lancaster, 2004). Furthermore, interviewing allows the researcher to creatively draw upon the interviewee’s perspective of the phenomena at hand, in a way which most other research methods cannot (Patton, 2002). The use of interviews as the primary data collection method was driven by the need to gain deeper insight into the phenomena than quantitative methods can offer (Patton, 2002).

Ten interviews were conducted in total, with the interviews lasting between 30 and 60 minutes. Seven interviews were conducted with helpdesk employees, two with project managers, and one with the Lead Developer of the development team. All interviews were conducted face-to-face at the department with the audio of the interviews recorded. Participants’ anonymity was guaranteed through careful transcription and coding of data to ensure any personal details or sensitive information remained confidential.

confident when approaching tasks?” Separate interviews were crafted for subordinates and managers to cover the phenomena from multiple perspectives. Both interview guides can be viewed in Appendix B and C.

Characteristics of Interviewees

Table 3: Characteristics of Interviewees

Along with interviews, observation was used to increase reliability. According to Kaplan and Maxwell (2005), observation provides the researcher with the ability to be able to draw inferences that allow us to understand, why an individual might hold a certain perspective or belief. Three hours of unstructured observations were conducted prior to the second and third rounds of interviews, and performed by watching the employees perform their daily routine while taking notes.

Data Analysis

Following data collection from semi-structured interviews and observation, audio recordings of the interviews was transcribed to enhance controllability (see Appendix E). Data gathered from participants was analysed in order to identify important themes and trends with regards to the aims of the study, which is to investigate the effects of

Interviewee Code Role Gender (M/F) Length of Time Worked in Current Role (y,m)

E1 Helpdesk Support M 0y, 2m

E2 Helpdesk Support F 0y, 9m

E3 Helpdesk Support M 2y, 10m

E4 Helpdesk Support M 1y, 1m

E5 Helpdesk Support F 1y, 6m

E6 Helpdesk Support F 1y, 1m

E7 Helpdesk Support M 2y

D1 Lead Developer M 6y

M1 Junior Project Manager M 3y

perceived transformational leadership on triggering Adaptive System Use behaviours. A secondary aim was to investigate whether perceived computer-self-efficacy has an effect on triggering these behaviours, or plays any other role in affecting perceived transformational leadership or Adaptive System Use.

To accomplish this this Eisenhardt’s (1989) directions on data analysis in qualitative research were followed, firstly conducting within-case analysis to become familiar with the content of each interview, followed by cross-case analysis to identify patterns across all the interviews. Open, axial, and selective coding of data was applied to identify concepts and relationships between codes (Corbin & Strauss, 1990). Deductive codes were created prior to the interviews and adapted from existing literature around Adaptive System Use, transformational leadership, and computer self-efficacy. Examples of deductive codes are: Intellectual Stimulation, Management Expectations, and Self-Leadership. Inductive coding was also performed, inductive codes are codes that emerge from the data and that do not (yet) appear in the existing conceptual models. Examples of inductive codes developed during the analysis are: Knowledge Transfer, and Simulation/Reproduction. Data that was linked to deductive codes was also re-examined following the inductive coding to assess its suitability for the new codes, allowing the researcher to assess the additional data to strengthen evidence on existing concepts as a kind of chain of evidence (Corbin & Strauss, 1990; Eisenhardt, 1989). Through axial coding, concepts and their dimensions as illustrated by the data were developed and presented in tables. Through selective coding, these concepts were connected into explanations of the behaviours and triggers of Adaptive System Use. The output of axial and selective coding is presented in the findings. Cross-case analysis such as this then allows the research to move toward generating new theories or conceptual models (Eisenhardt, 1989).

Quality Criteria for Research

Following van Aken et al. (2012), there are three quality criteria on which research can be assessed – controllability, reliability, and validity. All three aspects of these criteria are necessary is for inter-subjective agreement (van Aken et al., 2012, Swanborn, 1996).

FINDINGS

The findings of the effects of transformational leadership and computer self-efficacy on Adaptive System Use behaviours, in comparison with Sun’s (2012) model, and new findings, are presented here. The tables explain the representations of the Chains of Evidence presented.

The table below gives an overview of the respondent codes.

Adaptive System Use Behaviours from Sun (2012)

Data analysis revealed that from Sun’s (2012) model, of the four Adaptive System Use behaviours, trying new features and feature repurposing, were observed in the data, as illustrated in Table 6.

Table 6: Chain of Evidence, Current Adaptive System Use Behaviours

Respondent Code E1 E2 E3 E4 E5 E6 E7 D1 M1 M2 Trying New Features Feature Repurposing Colour Meaning

Evidence of Adaptive System Use Behaviour

Inconclusive Evidence of Adaptive System Use Behaviour No Evidence of Adaptive System Use Behaviour

Table 4: Identifiers of Adaptive System Use Behaviours

Table 5: Interview Codes

Code Role

E Employee

D Developer

Trying New Features & Feature Repurposing

The prevalence of ‘Trying new features’ was clear as illustrated below: “…what a user actually does with software is maybe 10% of what’s possible… But there are so many more capabilities, and possibilities… we’re always looking for new ways of making our work efficient, of finding new ways of solving problems…” (E3)

Feature Repurposing was also observed. Respondents stated that they felt the scope of the software they use was too broad to say they act outside of its intended purpose: “I don’t know if… if ‘fogbugz’, if what we’re focusing on has a really ‘intended purpose’. I already said it’s a very broad product….” (D1). One example of feature repurposing in this case was creation of the wiki using ‘fogbugz’: “We used the wiki system as a help system, a knowledge base for informing our users and… well because it’s a wiki it does work, but it was not really… not really specifically designed to do this.”(M1)

New Adaptive System Use Behaviours

Three new Adaptive System Use behaviours were observed. Seven participants saw Creating Workarounds as key in being able to satisfy users’ needs: “Sometimes a teacher wants to have a group within a group, but that’s not possible in Nestor, you can’t make sub-groups. So in that case … I’ll think of a workaround in which the teacher is able to make groups that only the people that are in a certain group can see. Those kind of workarounds, that’s really something that we’re always looking for.” (E1)

Software Combining is an adaptation of Sun’s (2012) feature combining. Due to the various pieces of software utilized by the department it was felt that Sun’s (2012) definition of Feature Combining was too limited, as respondents mentioned combining features across multiple platforms to complete tasks: “…one of the things that our department excels at is Table 7: Chain of Evidence, New Adaptive System Use Behaviours

Respondent Code E1 E2 E3 E4 E5 E6 E7 D1 M1 M2

Creating Workarounds

Software Combining

integrating pieces of software […], everybody uses a different system and then we piece it all together.” (E3)

Simulation and Reproduction is an Adaptive System Use behaviour that is facilitated by the availability of a Staging Environment; a replica of the real Nestor Environment that enables helpdesk supporters to reproduce issues faced by users without damaging the real environment. In this reproduced environment, they test solutions for the user’s problems, which they later apply to the real environment. Seven respondents mentioned simulation during interviews. Respondent E4 explains the motivation behind these behaviours:

“So if we see something unexpected there […] then we try to reproduce it on another server. And… well experimenting or trial-and-error approaches are indeed encouraged. We try to… just find it out, try everything to make sure you can find the failure and what causes it.” (E4)

Triggers of Adaptive System Use

This section will assess the presence of Sun’s triggers of Adaptive System Use before introducing new concepts. In line with Sun’s (2012) work, it was expected Novel Situations and Discrepancies trigger Adaptive System Use behaviours, and Deliberate Initiatives would not trigger Adaptive System Use behaviours. Table 8 explains the colour-coded responses presented in the tables.

Table 8: Identifiers of Potential Triggers of Adaptive System Use

Colour Meaning

Triggers from Sun (2012)

The table below illustrates the prevalence of Sun’s (2012) triggers of Adaptive System Use as analysed from the interviews.

Novel situations were supported by some evidence from the data analysis, with five respondents mentioning it as a trigger. One novel situation mentioned in the data was the example of how employees deal with new tasks: “How do I engage that software for that new task? Well I mainly click around first when it’s really new […] I just click around until I see some words that… look, like meaningful to me for completing the task.”(E4) Novel Situations often triggered Simulation/Reproduction and Trying New Features, as individuals sought to make sense of the task by experimenting with different features to advance towards solving the problem. Also noted here was a clear divide between respondents. Respondents who did mention novel situations as a trigger tended to be those who had a higher level of computer self-efficacy, and discussed in the interviews how they would use the software in a novel situation. Respondents with lower computer self-efficacy mentioned that they would assign tasks to someone else, or ask a colleague for advice:

“…we have one guy who’s really good with Syllabus so he’s going to give us a masterclass in order to help us understand Syllabus. Then what will happen next probably is if I see a case about Syllabus and I think this is the way to solve it but I’m not sure, I can just assign the case to him…” (E2)

The findings regarding Deliberate Initiatives strongly correlated with Sun’s work. All of those in senior roles and five employees dismissed the idea that their manager requests them to use the software in a certain way: “I think we always have room for experimentation, and I think also they encourage it, […] for example if there’s an error they always say “Yeah, find it you can reproduce it and experiment with it.” (E5)

Table 9: Chain of Evidence, Triggers of Adaptive System Use from Sun (2012)

Respondent Code E1 E2 E3 E4 E5 E6 E7 D1 M1 M2

Novel Situations

Finally, the findings for Discrepancies also lined up with Sun’s (2012) work. Sun posited that discrepancies were the strongest trigger for Adaptive System Use. Findings from this study reflected that Discrepancies triggered Creating Workarounds, as illustrated by this quote: “When a teacher asks us how he should create a certain situation within Nestor sometimes these situations are very context-based for his course. Sometimes even so specific we’re not allowed - or able - to create the kind of process or situation regarding groups, setups or anything else that he wants, so then we try to create a workaround so we get the effect he’s trying to create.” (E5) Discrepancies were also found to trigger Simulation/Reproduction behaviours: “Then we have the problems with the bugs and the things that don’t happen that regularly and then…. yeah, you first try to figure it out yourself, and if that doesn’t work you can go to staging and try to simulate the problem.”(E2)

Potential New Triggers of Adaptive System Use from Perceived Transformational Leadership

This study aims to examine effects of perceived transformational leadership on Adaptive System Use behaviours. Following the literature review and analysis of the data potential triggers were grouped into two categories, Primary triggers, and secondary triggers. Primary triggers are triggers that directly contribute to the performing of Adaptive System Use behaviours through perceived transformational leadership. Secondary triggers, meanwhile, are groundwork for Adaptive System Use behaviours, acting as a trigger for the triggers, but unable to directly trigger Adaptive System Use behaviours by themselves.

Primary Triggers - from Transformational Leadership Literature

Individualized Consideration was derived from transformational leadership literature from Bass (1999), the idea being that individualized consideration would inspire an Table 10: Chain of Evidence, Potential Primary Triggers of Adaptive System Use from Transformational Leadership – Deductive Codes

individual to perform Adaptive System Use behaviours. Findings however suggested that there was no correlation between Adaptive System Use and Individualized Consideration. This may be due to the coaching setup focusing upon the group level as the project manager illustrated: “I used to do that individually, and I moved – and that’s also because I’m too busy and the group is simply too big. I moved to, say, two times a year we have meetings where we talk about what they’re responsible for, how they perform with respect to the job and maybe even personally.” (M2)

Cho et al. (2011) described intellectual stimulation as the stimulation of individuals’ problem-solving skills through the challenging of their current perspectives with fresh inputs. Employees often mentioned the challenging of their own perspectives from management: “Different perspective… she really tries to teach you, especially from the way she thinks. Approach it differently, and try to show somebody why she’s approaching it from a different angle, and hope that we can take - not copy it - but take from it those things that are helpful and meaningful, and then combine it with our own to maybe come up with an even better solution” (E3).

Primary Triggers arising from the Data

The category of self-leadership and self-teaching was a new trigger crafted from Houghton, Dawley, and DiLiello (2012). Seven respondents suggested self-leadership and self-teaching was a trigger of Adaptive System Use behaviours within the department. The junior project manager explained how the department seeks to encourage self-leadership: “So… that’s for everything we do, we try to put them in charge, cause they’re the ones having to resolve the issues. Only for the more strategic decisions, the software that we support, for example, we take the final decision.” (M1). Self-Leadership was noted to trigger several Adaptive System Use behaviours, including: Creating Workarounds, Software Combining, Trying New Features and Simulation/Reproduction. Respondent E2 described their problem-solving process: “…you first try to figure it out yourself, and if that doesn’t work you can go to staging and try to simulate the problem, if that doesn’t work you can always ask a colleague.”

Autonomous Working is another inductive code arising from the data. Evidence suggests Autonomous Working positively influences Adaptive System Use, with eight respondents mentioning it as a trigger. Autonomous Working was found to trigger Creating Workarounds, Simulation/Reproduction, Feature Repurposing and Software Combining: “there are always several ways to find the information you need […] for example we have ‘Bombix’ which contains a lot of information about students, sometimes you can find the same information there which you can also find in Nestor […] there are several ways to find the same information and you’re absolutely allowed to use that…” (E2). Indeed, under the influence of this trigger, the lengths to which the employee is willing to perform Adaptive Table 11: Chain of Evidence, Potential Primary Triggers of Adaptive System Use from

Transformational Leadership – Inductive Codes

System Use behaviours is potentially only limited by their computer self-efficacy: “It’s like one big experiment, one big anarchy. The main rule is “don’t break the system,” and everybody has their feelings about what they can or can’t do within Nestor without breaking it.” (E4)

Finally, Creative Problem Solving is an inductive code arising from mentions of collaboration, experimentation, and creativity in solving problems in the helpdesk team. The Chain of Evidence regarding Creative Problem Solving, however, showed an inconclusive result. This could be a result of individual creativity being a function of computer self-efficacy, as one interviewee mentioned: “Sometimes when I know something about it then I feel confident, but if I don’t know a lot about it then I don’t really feel confident, and then I ask some of my colleagues who have a lot more knowledge about computers.” (E7)

Secondary Triggers of Adaptive System Use from Perceived Transformational Leadership This study also identified potential secondary triggers of Adaptive System Use. The Table below presents the Chain of Evidence for these triggers.

Two-Way Communication is an inductive code arising from the data, as respondents referred to the nature of the relationship between managers, themselves, and their colleagues. Respondents felt there was a two-way conversation going on as management communicated their objectives and expectations to the subordinates, while the subordinates provide feedback to the management. One respondent mentioned their observations from the meetings they attended: “…one thing I noticed in the meetings …there’s really a two-way conversation

going on. Like, if we point something out which we think “Ok this is not efficient,” or “this can be better” then I think they really listen to our input.”(E1)

Loss of knowledge is a key issue facing the department, with transfer of knowledge being heavily encouraged via documentation and word-of-mouth. Furthermore, Knowledge Transfer has properties conducive to activating first-level triggers of Adaptive System Use, such as facilitating self-leadership and self-teaching behaviours. Respondents mentioned the importance of sharing knowledge and its effects on individuals Adaptive System Use: “…it’s never the case that someone says “here you go, new software, figure out that user’s question.” They’ll always make sure that there’s at least someone who’s really experienced with that software, and can share that knowledge in order for you to gain more confidence… (E2).

Openness to Change was also identified as a potential secondary trigger of Adaptive System Use, however evidence proved to be inconclusive; only two employees mentioned openness to change as a trigger for two other primary triggers, intellectual stimulation, and management expectations: “So, your mindset as I could refer to, may be the best word. She’s changing your mindset into being more open towards changes or anything…” (E4). Looking at the whole group however, most did not share the feelings of the respondent above.

Articulating a vision is another aspect of transformational leadership derived from literature. It was thought articulating a vision of a future where both the employees and organizations interests are fulfilled increases employee commitment and drive (Bass, 1999). However, only three employees mentioned it as a trigger of Adaptive System Use. “Basically she’s very objective-driven as well, she always says “I want solutions, not workarounds.” So… yeah you’re always encouraged to be creative, and never to think of easy solutions but always think in the real long-term solutions” (E3).

The final potential secondary trigger refers to organizations culture. Openness and communication was a key part of the culture at the department, with management and developers working in tandem to attempt to foster an open culture, where “everyone is allowed to make mistakes, but, at least admit that they made a mistake and we can just have a talk about it” (M1). The data analysis determined that the organisational culture was an activating condition for all first-level triggers of Adaptive System Use.

Computer Self-Efficacy and Triggers of Adaptive System Use

arising from the data analysis, offer insights on the respondents existing levels of computer self-efficacy, and on their ability to engage in Adaptive System Use.

Information Overload refers to the individual having a large amount of information, or tasks to manage. The need to manage these affects individuals’ computer self-efficacy as they become distracted from fully utilizing the software’s features, damaging their ability to engage in Adaptive System Use. System complexity refers to the complexities of the multiple software that make up the work ‘system’ that employees use. These complexities limit individuals ability to act upon triggers transformational leadership-based of Adaptive System use in individuals with lower computer self-efficacy as they struggle to understand the system.

Software Limitations is self-explanatory. User’s computer self-efficacy determines how they respond to these limitations, individuals with higher self-efficacy use these limitations as a trigger to engage in Adaptive System Use, while individuals with lower self-efficacy may be discouraged from doing so. Comfort Level with Software, through computer self-efficacy plays a role in determining how users respond to triggers of Adaptive System Use. As with Software Limitations, comfort level with software can trigger or prevent engaging in Adaptive System Use.

questions, a lot of phone calls and sometimes we lose the overview…. That’s something we sometimes struggle with, especially when there are a few projects that have a limited number of support employees.”(M1)

Four employees suggested that complexity of the system was a barrier to Adaptive System Use, only the developer had a high enough level of computer-self-efficacy to overcome this barrier. System Complexity can prevent employees from using the software to its full capacity, as the developer acknowledged: “sometimes features are quite complex by themselves, sometimes you really need to have some kind of documentation to see what they mean with this small dialog box and these settings.” (D1)

Experiencing Software Limitations was perceived as nurturing computer self-efficacy, which in turn triggered Adaptive System Use behaviours as users utilized their skills overcome these limitations. Four employees mentioned that software limitations caused them to explore alternate ways of completing tasks: “…sometimes a teacher wants to have a group within a group, but that’s not possible in Nestor, you can’t make sub-groups. So in that case, if I get a question like that, I’ll think of a workaround in which the teacher is able to make groups that only the people that are in a certain group can see.”(E1). In some users, therefore, experiencing software limitations influences a user’s perceived computer self-efficacy, which in turn influences Adaptive System Use. Other respondents however, suggested they would take a more careful approach.

Comfort Level with Software yielded mixed findings, with users with higher computer self-efficacy more likely to leave their familiarity pocket and explore the software’s functionalities while users with lower computer self-efficacy felt less ready to do so, in line with Yamauchi & Swanson, (2010). Some respondents felt it may trigger Adaptive System Use, while some felt it may be a barrier. As respondent E3 mentioned, it may be dependent upon circumstance:“…when we have new software, that’s always a bit enthusiastic, see how it works, see what it does, and generally the first encounter will determine how you feel. Sometimes it has a really good look and feel, you know exactly where to look on your first go […] Sometimes you just see “Seven-thousand buttons, God help me.”(E3)

Computer Self-Efficacy as a Mediator between Transformational Leadership and Adaptive System Use Behaviours

DISCUSSION

Answering the Research Questions: What effects do Perceived Transformational Leadership and Computer Self-Efficacy have on an individuals’ Adaptive System Use?

This study has found evidence for two of Sun’s Adaptive System Use behaviours, namely trying new features and feature repurposing. Feature Substituting and Feature Combining were not observed. Feature Substituting was not found due to the nature of the software used by the department. The use of specialist software makes it difficult to substitute features as the software the department utilizes was developed with a specific purpose in mind. This software was adopted by the helpdesk because they already possess the sufficient features necessary for the helpdesk to perform its day-to-day activities without having to substitute features. Feature Combining was mutated into Software Combining following the data analysis as respondents mentioned combining multiple features from multiple platforms multiple software together at once, in real-time to solve problems, while Sun’s definition of Feature Combining merely referred to the use of features from other applications in conjunction with Microsoft Word (Sun, 2012, p.473).

The study identified three new Adaptive System Use behaviours that supplement Sun’s, namely Simulation/Reproduction, Creating Workarounds and the aforementioned Software Combining. Simulation/Reproduction is specific to the department studied due to the existence of the staging environment, without which these behaviours would not be possible. Further studies may wish to research whether Simulation/Reproduction is generalizable to other organizations that utilize staging or test environments. Creating Workarounds should be generalizable to multiple organizations, as the possibility to create workarounds exists within virtually every piece of software (Carzaniga, Gorla, Perine & Pezzé, 2015; Liu, Lin & Jiang, 2013; Mélard, 2014). Current Adaptive System Use triggers may also trigger these new behaviours; for example, Creating Workarounds may be triggered by Novel Situations, dependent upon receptivity to the trigger. Rogers (1995) notes innovative individuals are more likely to be receptive to new ideas that produce innovative behaviour.

Primary triggers directly trigger Adaptive System Use behaviour, as long as the individual has the ability and will to act upon the trigger as posited by Sun (2012), and has sufficient computer self-efficacy to be able to overcome the self-efficacy related barriers to Adaptive System Use.

The findings of this study that transformational leadership has an influence on IT users’ behaviour is in line with the work of Cho et al., (2011) who posited that transformational leadership positively influences information systems success, and other authors who examined these concepts. It refines the work of Cho and others by illustrating which specific dimensions of transformational leadership trigger which specific IT use behaviours. Of the four dimensions of transformational leadership behaviours conceptualized by Bass (1999): intellectual stimulation, inspirational motivation, individualized consideration and idealized influence, two trigger Adaptive System Use behaviours – intellectual stimulation and inspirational motivation (through the setting of expectations). These two dimensions were discovered to trigger creating workarounds, simulation/reproduction, and software combining behaviours.

Furthermore, it advanced the work of Jasperson et al., (2005) by answering their call for research on what influences users to utilize the full range of an application’s features during the post-adoptive stage of IT implementation. Additionally we expanded on work from Yamauchi and Swanson (2010), who examined the existence of individuals ‘familiarity pockets’ in the adoption phase by examining it the post-adoption phase, showing that these ‘familiarity pockets’ where users prefer to stay rather than venture outside of still exist in this phase of IT adoption.

In terms of the effects of computer efficacy, this study has found computer self-efficacy mediates the relationship between the triggers of Adaptive System Use and its outcomes, filling a research gap proposed by Sun (2012) and confirming Compeau and Higgins (1995) theory that computer self-efficacy has a mediating effect upon environmental factors, in this case transformational leadership.

Adaptive System Use behaviours, while individuals with weaker personal/cognitive factors struggled to respond to the interactions of the environmental influences (transformational leadership) with their own personal/cognitive factors (computer self-efficacy) and therefore were less capable to act upon the transformational leadership-based triggers to Adaptive System Use.

Secondary Triggers of Adaptive System Use

Primary Triggers of ASU from Transformational Leadership

Adaptive

System Use

Computer Self-Efficacy Moderators

Self-Leadership & Teaching Autonomous Working Intellectual Stimulation Management Expectations Two-Way Communication Knowledge Transfer Open Organizational Culture Revising Content of FIU Revising Spirit of FIU System/Software Related ASU Triggers Software Limitations

Software Complexity, Software Comfort Levels

H1 H2 H4 H5 H6 H7 H8 H3 H9ab H10ab