Empirical Analysis of the Effect of demographics; Age and Gender on
Transformational IT Leadership and Adaptive System Use
“Do Transformational IT Leadership, Age and Gender, Influence Adaptive System Use and
if so how?”
21 November 2016
Ahmed Özışık
S2549727
a.ozisik@student.rug.nl
Master ThesisMSc. Business Administration – Change Management Faculty of Economics and Business
University of Groningen
Supervisor:
dr. J.D. van der BijSecond supervisor:
dr. I. Maris-de Bresser2
Abstract
This research investigates the gap between the effects of transformational IT leadership, age and gender on adaptive system use (ASU). This is achieved by adding a forth trigger (transformational IT leadership) and two moderators (age and gender) to the conceptual model of Sun (2012). With the help of a questionnaire, 120 valid responses are retrieved; illustrating the perception of team members, and allowing an empirical research to be conducted. Findings show that transformational IT leaders motivate their users, and foster a creative environment, encouraging thinking of potential complications for a better collective outcome, and sprouting IT innovation for adoptive system use. But not in the sense that users are motivated to go above and beyond to use (new) IT. The effects of gender for the majority illustrated that for the relationships between the triggers and the dimensions of ASU, females either strengthen the relationship or weaken them, more than males, as expected. However, a major item to be considered is to not put all females in one box and all males in another so that transformational IT leaders could be misguided in their approach to effective use when a female for example does not fit the labeled stereotype. Findings for age suggested more mixed outcomes. One unexpected finding suggests that older people are less motivated then younger ones, to perform ASU. Findings insinuate that they are not game-changing and as significant as was expected, or for some, in the way the relationship was expected. Findings also insinuate that there could be many more (internal, external and situational) factors beside post-adaptive system use, still unexplored for this conceptual model to be beneficial for managers to use as a model for effective use and managing technological change.
3
Table of Content
Abstract ... 2
1. Introduction ... 4
2. Theoretical Framework ... 6
2.1 Adaptive System Use ... 6
2.2 Triggers of ASU ... 8
2.3 Contextual influences ... 16
3. Methodology ... 24
3.1 Sample and Data Collection... 24
3.2 Analysis ... 28
4. Results ... 30
5. Discussion and Conclusion ... 36
References ... 45
Appendix A – Conceptual Model ... 53
4
1. Introduction
Information Technology (IT) changed the way people work, both within and outside firm boundaries. Efficiently using IT is becoming an underlying part of a firm’s strategy in an increasing number of business organizations (Izquierdo, Samaniego, & Cabezudo, 2016), representing one of the most transformative impacts on business. Gartner Inc. (2016) predicts that corporate spending on IT will reach over $3.54 trillion in 2016. Despite these humongous IT (change) investments organizations do not meet the full benefits and potential out of them, as benefits do not arise from money spend on implementation, but on how people perform their roles in a more efficient and effective manner (Peppard, Ward, & Daniel, 2007). Thus, benefits emerge when people adapt to fit the system to their needs, making adaptive system use (ASU) an admissible strategic importance. The term ASU symbolizes a period of adaptation, where users actively revise their IS usage. These adaptations allow for exploitation and reaching the full potential of an IS, which enhances task performance (Tyre & Orlikowski, 1994; Mukhopadhyay, Kekre, & Kalathur, 1995; Jasperson, Carter, & Zmud, 2005).
Accordingly, several authors (Guines & Markus, 2009; Hsieh, Rai, & Xu, 2011; Nan, 2011) explored ASU, however, until recently, prior research (Ahuja & Thatcher, 2005; Barki, Titiah, & Boffo, 2007; Boudreau & Robey, 2005; Jasperson, Carter, & Zmud, 2005; Leonard-Barton, 1988; Saga & Zmud, 1994) focused on system level (IS acceptance, IS implementation, IS success, and IS-aided decision-making), whereas revision usually transpires at the post adoption phase (feature level), leaving an inaccurate description of how people revise their system use (Sun, 2012). This revision of system use is conceptualized by Sun (2012) in his research model as: Features In Use (FIU), which exists of trying new features, feature substituting (replacing features in FIU with other similar functions), feature combing (mixing with other features), and feature repurposing (using features other than what it was intended for).
According to Sun (2012), some of these features can be used voluntary, some less voluntary, and some without awareness. He investigated this area by creating the ASU model, with three new triggers (novel situations – experiencing unfamiliar things, discrepancies – outcomes of system use are different from the expectations, and deliberate initiative – one is requested to revise use of system features) that lead to ASU. With the last trigger – deliberate initiatives – leadership literature in the context of adaptive system is only specified for transactional leader theories, which pay attention on the (requesting) completion of orders and transactions (Bass, 1985; Burns, 1978).
As a consequence, an unexplored trigger is using through inspiration, a push that allows the user to inspire them to go above and beyond on how to use IT strategically. This vision strongly relates to the transformational leadership style and is needed to advance these features, adoption behaviors and innovativeness in IT, as transformational leaders foster an environment of creativity (Amabile, Conti, Coon, Lazenby, & Herron, 1996; Mumford, Scott, Gaddis, & Strange, 2002). Else way, innovation cannot occur. As Ronald Reagan, the 40th president of the United States mentions; "The
5
people to do the greatest things". This inspiration to go beyond, described by Shamir (1991) as leaders
who are able to enthuse and motivate users by addressing common sense to the work, challenges employees to think beyond their mind-sets, resulting in innovation (Bass & Avolio, 1997).
However, the impact of the demographics; age and gender, usually gives different outcomes for males and females (Beel, Langer, Nürnberger, & Genzmehr, 2013). Also, older people tend to be more motivated in the fear of not losing their jobs (Gallie, Felstead, Green, & Inanc, 2016), while younger users are for example more experienced growing up with IT (Czaja & Sharit, 1993; Kanfer & Akkerman, 2004; Warr & Pennington, 1993). With this, a broader, more detailed view is taken, which is expected to allow for a more fine-grained result regarding peoples post-adoptive behaviours when age and gender are included.
With this in mind, transformational leadership can be specified further, as the concept of ASU deals with systems, in other words; technology, it points to the field of IT. Rucker (2013) explains the reason behind the specific need for transformational IT leadership, when trying to engage people with IT; "The pace of change in technology means that you always have some segment of your staff that
wants to learn the hottest new tools, and you need to keep them out front so that they stay engaged”.
Therefore, this inspiration to go above and beyond, combined with the aforementioned goal of ASU – enhancing task performance –, and fused with the humongous investments for IT in mind – to use it effectively –, leads this research to investigate transformational IT leadership as a new trigger of ASU. To shed light into this dark spot, this research aims to investigate the effects of transformational IT leadership in relation to users revising their ASU. In other words, it is important to understand and know how firms (transformational IT leaders) can motivate and with that enable their users to move in an adoptive direction, because in turn, this enables the benefits from IT. Combining these research directions, flowing from the gaps in literature, the following research question is formulated: Does Transformational IT Leadership, Age and gender, Influence Adaptive System Use
and if so how?
By answering this research question, it is intended to contribute to post-adoptive behaviours, gender, age, and transformational leadership. It is anticipated that this is realized by adding post-adoptive behaviour theories by restating the concept of ASU (Sun, 2012) under a different setting, and at the same time referring to the use of IT features in general instead of the specific system features that Sun (2012) proposed. Furthermore, it aims to contribute to transformational leadership theories by proposing the notion of transformational IT leadership with regard to ASU.
Together with age and gender, this research wants to investigate whether and if so, according to which (gender) employees’ perception (and age), the occurrence of transformational IT leadership is causing a change in their feature revisions.
6 retained surveys. These surveys investigate how and if transformational IT leaders can influence team members to use IT in an adoptive (innovative) manner. Next, results that are published through SPSS are elaborated. The discussion and conclusions are attained thereafter. Lastly, the conceptual model, surveys, and SPSS output are enclosed in the appendix.
2. Theoretical Framework
The theoretical framework examines the ASU model of Sun (2012) further, by looking at how transformational IT leadership as a new trigger can influence ASU, and how age and gender influence the triggers as moderators. This revised conceptual model (Appendix A) will shortly be elaborated, as this will highlight where this research seeks to add to existing literature. The reason for leaving the unsupported hypothesis is users’ different behaviors when combined with other triggers (Transformational IT Leadership) and moderators (age and gender), as authors (Beasudry & Pinnsonnealt, 2005; Jasperson et al., 2005; Leanoard-Barton, 1988) explain that users go through multiple adaption sequences, encountering multiple different types of triggers during a single adaptation cycle. Also removing an unsupported hypothesis from a research could make something significant, insignificant, and vice versa. This new trigger and moderator allows for eleven new hypothesesto be tested. Without further do, the theoretical framework continues examining ASU on a deeper level.
2.1 Adaptive System Use
Sun (2012), in order to understand how people revise their use of system features, developed a new concept of post-adoptive system use at the feature level, called adaptive system use. It is based on Louis and Sutton’s (1991) study on how people switch to active thinking from automatic thinking, including conditions that provoke these. Subsequently, Sun (2012) defined ASU as the user’s revisions of what and how features are used. It exists of combinations of individuals’ behaviors, which collectively describe how they actively revise their system use. The system features correspond to user’s tasks that the IS supports, and grouped by Harrison and Datta (2007) as feature groups. The revision of these system features is conceptualized by Sun (2012) as the basket of system features that are ready to be used by a particular user to accomplish tasks, called Feature In Use (FIU), and together with features from a variety of systems, called ecosystem (Sun, 2012).
7 A similar concept, familiarity pocket, is developed by Yamauchi and Swanson (2010). Here, personal work routines and the components are accumulated through situated interactive use of the system. It focuses on the gain through routines by which users interact with each other and the system. Accordingly, users can share routines with it through their interactions, and because each must interpret the actions of others, their respective familiarities with the routine will be different. This illustrates users’ acknowledgement regarding interactive systems use.
This works as follows; users firstly incorporate (and portray) moves from which routines may be composed, which is a compulsory condition for routine formation (Pentland, Feldman, & Becker, 2009). Secondly, each move is understood as interactive, linking users to systems. This interaction allows for feedback, motivating another move that leads to adaptive learning. Lastly, users occasionally make moves outside their familiarity pocket, which is usually incorporated into their own familiarity pocket. This concludes that routines are needed and serve as the basis for improvisation (Feldman & Pentland, 2003), and that over time user may use different features to cope with changing work and technical environments (Sun, 2012), possible from other users’ behaviours (Yamauchi & Swanson, 2010).
Accordingly, DeSanctis and Poole (1994), and Orlikowski (2000) suggest that these features and behaviors of ASU can be divided into two dimensions: revising the content of features in use and
revising the spirit of features in use. These dimensions are discussed next.
Revising the Content of FIU
The dimension, revising the content of FIU, is the user’s revision regarding what features he/she uses. Barki, Titiah, and Boffo (2007) showed that users revise the content of FIU by trying new features, and propose they belong to an individuals’ exploration behaviour, which often occurs. Here users independently seek a way to gain more knowledge regarding an IT (Saga & Zmud, 1994), and as they gain experience, they feel a wide variety is needed (Hiltz & Turoff, 1981). This exploratory behaviour (trying new feature), enhances ones’ knowledge of system features.
8 Revising the Spirit of FIU
Revising the Spirit of FIU, is the corresponding subdimension, where user’s revisions pertain how features are used. Harisson and Data (2007) highlight the meaning as the ways that allow them to best finish their tasks. One way is divided by Desouza, Awazu, and Ramaprasad (2007) as feature
combining, where features in ones’ familiarity pocket are mixed and matched (Rice & Rogers, 1980),
combining present existing features to reinvent (Rice & Rogers, 1980) new functionality that are used
in combination for the first time (Boudreau & Robey, 2005). The other way (subdimension) is by feature repurposing, where features are applied by users in different ways then they were invented for,
by the designers (Jasperson, Carter, & Zmud, 2005), extending the product to a new task (Singletary, Akbulut, & Houston, 2002). What also leaves to innovative ways to use IS effectively (Ahuja & Tatcher, 2005), is when users use add-ons to create workarounds (Desouza, Awazu, & Ramaprasad, 2007).
However, not all features are revisable (Sun, 2012), as they either contain (system) restrictions, or users are not aware of them. Thus, not all ASU behaviours are performed simultaneous- or identically and therefore can cause different extents. In addition, as people are neither always positive nor right about their effective use of system features (Abrahamson, 1991), they use active cognitive processing to create innovative behaviours like ASU (Sun, 2012). The next section therefore, explores the triggers, as these are seen to cause users to engage in this particular way of active thinking. Before doing so, the figure below summarizes the dimensions of ASU, as adopted from Sun (2012).
Figure 1
Dimensions and Subdimensions of Adaptive System Use
ASU
Dimensions of ASU Subdimensions of ASU
Revising the Content of FIU
User’s revisions of what features are included and used in their FIU.
Revising the Spirit of FIU
User’s revisions regarding how features in their FIU are used.
Feature Combining (FC)
Using features in FIU together for the first time.
Feature Repurposing (FR)
Using features in one´s FIU I in a new way.
Trying New Features (TR)
Add new features to one’s FIU and thus expanding the scope of the FIU.
Feature Substituting (FS)
Replacing features in the FIU with other features with similar functions.
2.2 Triggers of ASU
9 different from the expectations, and deliberate initiative – one is asked to revise use of system features) which he subtracted from Louis and Sutton’s (1991) study.
The new framework created for this study (Appendix A) depicts an ASU episode, containing multiple adaptation sequences (Beaudry & Pinsonnealt, 2005), and where outcomes through feedback loops (Pearlson & Saunders, 2010), and attempts of users themselves, are evaluated by them, can trigger another adaptation (Jasperson, Carter, & Zmud, 2005). Before diving in the triggers individually, it is important to note, that this ASU episode’s main learning method is through trial and error. Also, these triggers are believed to promote active thinking, not behaviors essentially (Sun, 2012). However, as Jasperson, Carter, and Zmud (2005) illustrated, they are compulsory criteria in order to obtain active behavior usage. In this sense, triggers are embedded in either contradictions or interruptions (Sun, 2012).
Novel Situations and its (Sub-) Triggers
These contradictions refer as the misfit (Soh & Sia, 2005) within (elements of) activities, manifesting themselves as issues, ruptures, breakdowns and clashes (Kuutti, 1995). Thus, making triggers like novel situations, manifestations of different types of contradictions (Sun, 2012). Novel situations are referred as the contradictions between a person’s current system and their new one, that user view as ‘out of the ordinary’ (Armstrong & Hardgrave, 2007), or as Sun describes it: “one is experiencing
unfamiliar things” (Sun, 2012; 460). Sun divided these unfamiliar experiences into three sub-triggers
from different authors, namely: when the user has to perform an unfamiliar task (new task) (Ahuja & Thatcher, 2005; Jasperson, Carter, & Zmud, 2005), when one is observing another’s system use (other’s use) (Compeau & Higgins, 1995; Boudreau & Robey, 2005; Ryu, Kim, Chaudhury, & Rao, 2005), and lastly, when one’s system environment changes (through revision of hard- and software), which Shaw (2001) and, Benamati, Lederer, and Singh (1997) portrayed. Accordingly, the following hypothesis is adopted from Sun (2012);
H1: Novel situations – in the form of new tasks, others’ use and changes in system environments – are positively associated with ASU.
Discrepancies
10
H2: Discrepancies are positively associated with adaptive system use.
Deliberate Initiative
The third trigger, deliberate initiative, occurs when someone is either explicitly requested to use or revise known features (Venkatesch & Davis, 2000), or inquired to try new system features (Sun, 2012). It is the initiative one takes in response to a request for an increased level of attention, when asked to think or while being explicitly questioned (Langer, 1986; 7). Examples of cases discussed (Benamati, Lederer, & Singh, 1997), portrays a conversation where a supervisor or boss requests the user to use certain features of a system of which the user is not familiar with. In other words, it suggests deliberate initiatives are positively associated with ASU, as they voluntarily or not, generates users to perform this third trigger, hence the hypothesis;
H3: Deliberate initiatives are positively associated with adaptive system use.
Although multiple authors (Jasperson, Carter, & Zmud, 2005; Langer, 1986; Louis & Sutton, 1991) believe that these abovementioned triggers give rise to ASU through active cognitive processing, Louis and Sutton (1991) elaborate that a single trigger can also be altered into a new one. Sun (2012) therefore, continues explaining that next to a direct impact on ASU, deliberate initiatives and novel situations can also exert an indirect influence on ASU through provoking discrepancies. This moderator is also consistent with Jasperson, Carter, and Zmud’s (2005) research of post-adoption system use, where people through the technology sense-making process, make sense of the situation. It associates with the reflection of the user’s personal system use, which could allow ASU through disconfirmation, a notion comparable with discrepancies. More specifically, through the feedback mechanism (Beudry & Pinsonneault, 2005; Jasperson, Carter, & Zmud, 2005) as ASU is a process that starts (mainly by trial and error) in a particular state and return to that state after they undertook a few different processes.
Novel situations can prompt one to start this trial and error. Here, early on, people tend to apply their current schemas of their system feature use, to new situations. In which case they may be exposed to realize that their current schemas are inadequate to deal with the renewed situation. This output judgement, combined with feedback from others, or the task itself, may generate discrepancies as expectations are not met. This process than renews for another trial and error process as novel situations occur more frequently. Benamati, Lederer, and Singh (1997) complement this by displaying that with a new in system implementation, users more often exhibit discrepancies when confronted with novel situations. Flowing from this thought of reasoning the following hypothesis is formulated;
11 This same line of reasoning applies to deliberative initiatives, where requesting a change in a users’ system use, can results in trial and error when outcomes are dissatisfying, leading to discrepancies, and the following hypothesis;
H5: Deliberate initiatives are positively associated with discrepancies.
However, being asked to change one’s system use may force a user to employ ASU behaviours with their current schema of system use (Schön, 1983). Schön elaborates that when confronted with demands, a person may demonstrate active thinking and behavior. These may result in unsatisfactory outcomes to other users, which can lead to discrepancies; hence the reasoning for the new trigger and hypotheses. Building on that notion, Hastie (1984) and Schön (1983) argue that people react to external causes, which can also act as antecedents for active thinking. Users may perceive this as (politely) forced or not obligatory as Agarwal and Prasad (1994) describes it. Hartwick and Barki (1994) share the same reasoning as they indicate that although there is an organizational requirement of mandatory use, the intention of a usage can vary because of a user´s individual willingness to adapt to new situations. To be able to investigate this notion, the next section looks in the opposite direction of deliberate initiatives, or in other words, another cause of FIU.
Transactional and Transformational Leadership
Following up, some of these FIU are selected voluntarily (e.g. tracking changes in Word), some automatically (automatic line break), some less voluntarily (save file), and some without awareness (word count). As aforementioned, Sun (2012) investigated this area by creating the ASU model, with three triggers (novel situations – experiencing unfamiliar things, discrepancies – outcomes of system use are different from the expectations, and deliberate initiative – one is requested to revise use of system features) that lead to ASU. With the last trigger – deliberate initiatives – leadership literature in the context of adaptive system is only specified for transactional leader theories, which pay attention on the (requesting) completion of orders and transactions (Bass, 1985; Burns, 1978).
12 Sun (2012) has already tried to investigate this with deliberate initiative. However, he investigated – a request directed to specifically change ASU –, a leadership style, strongly related to transactional leaders. The nuance with transformational leadership for ASU lies in the literature, which suggests transactional leaders are less effective when it comes to change, since they disregard anything that is close in relationship to any kind of interpersonal or organizational concern (Afshari, Bakar, & Luan, 2009). With the change focus on ASU in mind, it is therefore relevant to also investigate transformational leadership, as they enable followers to acquire the knowledge and skills they need in order to work more effectively, and accordingly perform better (Yukl, 2012). This is realized by inspiring users to go above and beyond (Andressen, Konradt, & Neck, 2012), attempting to motivate the individual motivational factor (Burke et al. 2006), in order to reach full potential (Dvir, Eden, Avolio & Shamir, 2002).
This braids on the notion mentioned in the introduction that benefits do not arise from money spend on implementation, but on how people perform their roles in a more efficient and effective manner (Peppard, Ward, & Daniel, 2007). Thus, only when people adapt to fit the system to their needs, do benefits emerge, making not only ASU an admissible strategic importance, but transformational leadership and IT (system) usage as well.
Furthermore, Jasperson, Carter, & Zmud (2005), and Bassellier, Benbasat and Reich’s (2003) have not yet explicitly addressed what influence transformational leaders have in order to stimulate behavior that investigates how the interaction of individuals with IT can be enhanced or stimulated, referring to the (post-) adoptive phase. This is important because this phase especially, is where the longest and most benefits is accrued for the firm (Jasperson, Carter, & Zmud, 2005), providing this research with another important reason to investigate this leadership style within ASU.
Furthermore, transformational leadership bears its importance within this research as Ronald Reagan, the 40th president of the United States mentions; "The greatest leader is not necessarily the
one who does the greatest things. He is the one that gets the people to do the greatest things". This
inspiration to go above and beyond, is described by Shamir (1991) as leaders who are able to enthuse and motivate users by addressing common sense to the work, challenging employees to think beyond their mind-sets, resulting in innovation (Bass & Avolio, 1997). This is needed in order for people to perform their roles in a more efficient and effective manner (Peppard, Ward, & Daniel, 2007).
Transformational IT leadership
With this in mind, transformational leadership can be specified further, as the concept of ASU deals with systems, in other words; technology, it points to the field of IT. Rucker (2013) explains the reason behind the specific need for transformational IT leadership, when trying to engage people with IT;
13 As a consequence, an un unexplored trigger is usage that relates to transformational IT leadership, a push that allows the user to inspire them to go above and beyond on how to use IT strategically (Andressen, Konradt, & Neck, 2012). This strongly relates to the vision of transformational leaders, which is needed to advance these features, adoption behaviors and innovativeness in IT, as transactional leaders foster an environment of creativity (Amabile, Conti, Coon, Lazenby, & Herron, 1996; Mumford, Scott, Gaddis, & Strange, 2002). Else way, innovation is difficult to occur, which Bass and Avolio (1997) emphasize is needed for employees to think beyond their mind-sets to work effectively.
Building on that notion, according to Thite (2000), there are essential characteristics of transformational leadership that are present in the behavior of successful IT project managers. Managers who exhibit certain leadership behaviors can improve IT projects (Andreu & Ciborra, 1996) and firms with high-level leadership skills are profound to sustain a competitive advantage over IT (Dehning & Stratopoulos, 2003). Therefore, it is presumed here that leaders who possess transformational leadership characteristics are able to motivate others to get the best performance when they are working with IT.
Their motivational inspiration to go above and beyond, combined with the aforementioned goal of ASU – enhancing task performance –, and fused with Gartner’s (2016) $3.54 trillion in corporate IT spending predictions for 2016 – to use it effectively –, leads this research to investigate transformational IT leadership as the fourth trigger of ASU.
This crucial link to overall effectiveness due to the humongous investments, evolves as benefits do not arise from money spend on implementation, but on how people perform their roles in a more efficient and effective manner (Peppard, Ward, & Daniel, 2007). Accordingly, Davis, Bagozzi, and Warshaw (1989) explain that it depends on the user’s intention to IT use. This later part is where transformational IT leadership emphasizes its utmost importance, as their characteristics according to several authors (Shamir, 1991; Bass & Avolio, 1997;) are more suitable to affect employees’ intentions positively as no other leader ship style (Hamstra, Van Yperen, Wisse, & Sassenberg, 2014), to get the best IT-usage effort from them (Bass, 1985, 1999).
Definition and reasoning
Accordingly, this research acknowledges transformational IT leadership as transformational leaders who foster and encourage IT usage. From this line of thinking, Bass (1985, 1999) defines IT leadership as a process of inspiring subordinates to share and pursue the leaders’ vision concerning IT and motivating others to move beyond their own self-interests in IT and work for the aims of the team by the use of IT.
14 transformational IT leadership in six dimensions, naming each six with a definition. Biernath (2014) and Sietsma (2014) revised these six definitions for IT, by adding small adjustments that narrows the transformational leadership style to the field of IT (Table 1).
Table 1
Dimensions of transformational (IT) leadership (Podsakoff, 1990; Biernath (2014); Sietse, 2014)
Dimension Transformational IT Leadership
Articulating a vision Behaviour on the part of the leader aimed at identifying new IT opportunities for his or her unit/division/company, and developing, articulating, and inspiring others with his or her vision of the feature in the field of IT.
Proving an appropriate model
How a leader behaves with IT sets an example for employees to follow that is consistent with the values in the area of IT that the leader espouses.
Fostering acceptance at group goals
Behaviour of the leader aimed at promoting cooperation through the use of IT among employees and getting them to work toward a common goal with the help of IT. High performance
expectations
The leader’s expectations for excellence, quality, and/or high performance in the use of IT on the part of followers.
Providing individualized support
Behaviour on the part of the leader that shows that he/she respects followers and is concerned about their personal feelings and needs (remained unchanged).
Intellectual stimulation Behaviour on the part of the leader that challenges followers to re-examine some of their assumptions that they have about IT and rethink these assumptions.
Although others have defined transformational leadership, transformational IT leadership is still a relatively new term for which authors use transformational leadership definitions, self- embedded with the corresponding (IT) field they research. The peer reviewed articles of Biernath (2014) and Sietsma (2014) regarding (the definitions of) transformational IT leadership allowed for more finer-grained area within transformational IT leadership, to be investigated. It contributed to the questionnaire transformed for this research, as the questions were formulated for the dimension of transformational IT leadership, which could be combined with the existing questionnaire of Sun (2012) for ASU. This finer-grained picture is anticipated due to the expanding view of the relationship between the dimensions of the trigger transformational IT leadership, and the dimensions of ASU in order to see if any, and if so in which piece what exact relationship is found for transformational IT leadership.
15 transformational IT leaders transpire overall job satisfaction (Andressen, Konradt, & Neck, 2012; Ford, Ledbetter, & Roberts, 1996), it is anticipated that intrinsic motivation could build on the eager search for effective use (Burton-Jones & Grange, 2012; Staehr, Shanks, & Seddon, 2012; Wagner, Newell, & Piccoli, 2010) in literature. Flowing from this line of thinking, the forth trigger, transformational IT leadership, is hypothesized as following;
H6: Transformational IT Leadership is positively associated with adaptive system use.
By combining the definition of transformational IT leadership with the abovementioned (sub) triggers, the following conceptualization is formed;
Table 2
Types of (Sub)Triggers (Sun, 2012)
Definitions of Triggers (Sub)Trigger Examples in System Use Novel Situations – are situations where a
person encounters things that are unfamiliar, previously unknown, unique, or that appear to be out of the ordinary.
- New task – the user experiences unfamiliar features. - Other’s use – one observes other’s system use.
- Changes in system environment – a firm’s system (hardware, software) is upgraded.
Discrepancies – represent situations where unexpected failures, disruptions, or significant differences exist between expectations and reality.
- Unexpected system feature failures.
- Outcome of system usage is different from what is expected. - One is requested to use certain unfamiliar features, changing system environments.
Deliberate initiatives – are the initiatives one takes in response to a request for an increased level of attention, when requested to think, or while being explicitly questioned.
- A user is requested by his or her supervisor to revise his or her system with (unfamiliar) system features.
Transformational IT leadership – is the process of inspiring subordinates to share and pursue the leader’s vision concerning IT and motivating others to move above and beyond their own self-interests in IT.
- A user is inspired by his or her leader to seek and use system features with which he or she is not familiar with.
- A initiatives one takes in response to a leaders’ inspiration, where one is inspired to revise use of system features.
However, these triggers are not solely the cause of the differences in outcome and behavior. A person’s age and gender (Beel, Langer, Nürnberger, & Genzmehr, 2013), can also play a huge role in her/his thinking and behavior process, next to Facilitating Conditions and Personal Innovativeness of
IT, which Sun (2012) investigated. As these (demographics) moderators (also called contextual
16
2.3 Contextual influences
These (demographic) moderators can be divided into internal and external influence types, which provide a more in-depth classification of the moderator and the reason they occur. Influence types can be explained by building on Louis and Sutton’s (1991) point, that the presence of triggers does not assure active thinking and behavior, but that they are depended upon individual and contextual factors. Because when someone encounters a trigger, they need to have the ability and be willing to notice it (Burke et al., 2006; Langer, 1986), hence individual (internal) factors. External context on the other hand, refer to contextual factors external to a person, such as the external support users receive from their environment. Now that contextual influences are split up, the next section divides the moderators accordingly.
Facilitating conditions
Sun (2012) building on Louis and Sutton’s (1991) abovementioned argument, categorises facilitating conditions as an external contextual influence factor, due to its definition: the degree to which an
individual believes that an organizational and technical infrastructure exists to support his or her use of a system (Venkatesch, Morris, Davis, & Davis, 2003; 453). This is explained by the notion that
users often look for external support when innovating (Scott and Bruce, 1994; Tierney, Farmer, Graen, 1999), to remove barriers of system usage, which observers believe is to support finishing their task (Venkatesh, Brown, Maruping, & Bala, 2008).
Moreover, facilitating conditions are closely related to how much control users perceive of what they are doing (Azjen, 1985, 1991; Taylor & Todd, 1995; Venkatesh, Morris, Davis, & Davis, 2003). As users can perceive novel situations of new IT with frustration and worries (Morris & Venkantesh, 2010), the necessary support and resources that facilitating conditions provide, can allow users to overcome uncertainties by increasing the probability of success that users calculate before taking action (Venkatesh, Brown, Maruping, & Bala, 2008). This increased probability helps users to overcome ambiguities, which could help in making behaviours more controllable and achievable, which in turn makes users more likely to respond to novel situations (Sun, 2012). Thus, because facilitating conditions is believed to reduce the frustration and worries (Basowitz, Persky, Korchin, & Grinkler, 1955), it is anticipated that it also reduces these possible negativities between the relationship of novel situations and the dimensions of ASU, leaving the following hypothesis behind;
H7a: The effect of novel situations on ASU will be moderated by facilitating conditions such that this effect will be stronger when facilitating conditions are sufficient than when they are scarce.
17 leverage this support by exploring more discrepancies and with that increasing the likelihood of taking action to respond to unknown situations (Sun, 2012).
However, while with novel situations users spend their own work time to learn the system, with discrepancies, users are expected to spend their own additional time on top of their work time, making it possible for users to perceive that they are overloaded. This may even lead to anxiety or burnout (Jackson et al., 1987). Facilitating conditions can lighten these overloaded feelings users encounter with discrepancies, by providing them with access to resources, knowledge and assistance needed to perform their task (Sun, 2012; Venkatesh, Brown, Maruping, & Bala, 2008), that otherwise would not be possible based on the skills of the user alone (Ahuja & Thatcher, 2005; Sales, 1970).
In other words, facilitating conditions is believed to reduce negative reactions (Sun, 2012) between the relationship of discrepancies and ASU by providing support (by reducing barriers), which is expected to be beneficial to users to more likely perform ASU behaviours. Hence Sun’s (2012) following hypothesis;
H7b: The effect of discrepancies on ASU will be moderated by facilitating conditions such
that this effect will be stronger when facilitating conditions are sufficient than when they are scarce.
Autonomy is also beneficial for users, as it is utmost important for innovative ASU behaviours (Feist, 1999; Feist & Gorman, 1998) because several authors (Ahuja & Tactcher, 2005; Breaugh, 1985) illustrated that autonomy positively influences a person effort to innovate with IT. However, therefore users need to decide how, and at what pace ASU behaviours are performed (Ahuja & Tatcher, 2005; Breaugh, 1985). Deliberate initiatives - where one is asked to use specific features - may be perceived as controlling, and can give rise to perceptions of constantly watched (Barnowe 1975; Deci, 1975; Ryan, 1982; Ryan & Grolnick, 1986; Scott & Bruce, 1994). This can result in an unpleasant autonomous climate, where users could resist deliberate initiatives.
In addition, when users are asked to use certain features differently, they could experience a loss in intention to use them (Saeed, Abdinnour, Lengnick-Hall, & Legnick-Hall, 2010). This unpleasantness can weaken the relationship between deliberate initiative and ASU.
Allowing users to be more flexible in adaptation (e.g. by letting them chose more options and its pace), can broaden their autonomy. This helps users to perform more freely within the same timeframe, allowing facilitating conditions to help soften potential negativity between the relationship of deliberate initiative and ASU and with this, presenting Sun (2012) his following hypothesis;
18 While, transformational IT leaders foster and encourage IT usage (Amabile, Conti, Coon, Lazenby, & Herron, 1996), they still are a huge factor in effecting the perception users have to innovate and use features (Wand & Walumbwa, 2007). This depends on how good users are motivated by transformational IT leader to explore and innovate (Agarwal, 2004). Facilitating conditions can influence users’ perceptions that they are being better managed by adding adaptations (e.g. giving more support to users). Kates, Travis, and Wilbanks (2012) argue that these (support) actions than becomes transformational. In other words, when these adaptions are integrated with the approach of a transformational IT leader, it can result in in a much larger effect than before (Kates, Travis, & Wilbanks, 2012). Tierney and Farmer (2004) complement this by arguing that users perceive creativity as being wanted when leaders are more supportive, hence the believe that these adaptations of facilitating conditions could ease the relationship between transformational IT leadership and ASU. From this line of thinking flows the following hypothesis;
H7d: The effect of transformational IT leadership on ASU will be moderated by facilitating conditions such that this effect will be stronger when facilitating conditions are sufficient than when they are scarce.
Personal Innovativeness in IT
Sun (2012) categorised personal innovativeness in IT (PIIT) as an internal contextual influence factor due to its definition and because ASU is by nature already innovative. While the definition of PIIT is kept as it already resembles the category itself, another reason is to compare it with Sun’s outcome for which the meaning requires to be identical. Furthermore, by staying true to this definition, the research model and the survey questions, could remain unchanged, preventing inconsistencies of making items significant insignificant, and vice versa. Agarwal and Karahanna (2000), and Agarwal and Prasad (1999) defined PIIT as an individual trait reflecting one’s willingness to experiment with new IT.
The aforementioned triggers, however, do not always show themselves as triggers, meaning a user needs to be able recognize the triggers in order to be innovative (Burke et al., 2006; Langer, 1986). Zhou (2003) complements this by illustrating that users differ in levels of sensitivity to new and/or creative ideas, and so in the ability to produce innovative outcomes. In addition, innovative users are more likely to perceive new information through higher PIIT, needed for innovative behaviour (Rogers, 1995), to develop more positive perceptions of IT innovation than others (Agarwal & Prasad, 1999). Consequently, a user with high PIIT is expected to have a higher chance to perceive novel situations and discrepancies (Sun, 2012).
19 (Kirton, 1976), and are more likely to accept these risks and uncertainties to engage with features when encountering triggers. Therefore, it is presumed that PIIT could have a positive effect on the relationship between novel situations and ASU, and discrepancies and ASU. Hence the following hypotheses:
H8b: The effect of novel situations on ASU will be moderated by personal innovativeness in IT such that this effect will be stronger for individuals with high personal innovativeness in IT.
H8c: The effect of discrepancies on ASU will be moderated by personal innovativeness in IT such that this effect will be stronger for individuals with high personal innovativeness in IT.
Continuing, the impact of deliberate initiatives on ASU is, unlike novel situations and discrepancies, expected to have a negative moderating effect. This is because less innovative (low PIIT) users, are more uncertain and therefore more likely to be influenced by behaviors and follow guidance from others (Zhou, 2003), making innovative people in turn likely less responsive to external requests (Sun, 2012), as they foster high autonomy, self-confidence, and flexibility, as characteristics, to follow orders (Feist, 1999; Feist & Gorman, 1998). Similarly, the resisting controlling behaviour of feeling continuous watched, which was mentioned by deliberate initiatives before, could conflict with users’ strong orientation towards their autonomy (Deci & Ryan, 1987; Feist, 1999; Feist & Gorman, 1998; Greenberg, 1992; Oldham & Cummings, 1996) if their PIIT is high.
To conclude, a request for users to revise system features can cause resistance by challenging their autonomy, insinuating that users with high PIIT, can have a negative effect for the relationship between deliberate initiative and ASU, building the hypothesis (Sun, 2012);
H8d: The effect of deliberate initiatives on ASU will be moderated by personal innovativeness of IT such that this effect will be weaker for individuals with high personal innovativeness in IT.
20 one’s leader. With that, users might view their transformational IT leader as visionary, which can have a positive effect on the inspiration of users (Dvir, Eden, Avolio, & Shamir, 2002) with high PIIT. Similarly, having a comparable perception and view as their transformational IT leader, could be positively perceived by users (Felve & Schyns, 2010) with high PIIT, and in turn could improve users’ perception of their leader. Therefore, it is expected that for users with high PIIT, the relationship between transformational IT leadership and ASU could be eased and received as positively, allowing for the following hypothesis to be formulated;
H8a: The effect of transformational IT leadership on ASU will be moderated by personal innovativeness in IT such that this effect will be stronger for individuals with high personal innovativeness in IT.
Gender
A person’s gender (Beel, Langer, Nürnberger, & Genzmehr, 2013) also plays an important role in one’s thinking and behavior process. This demographic moderator is an internal contextual influence factor, as it is linked to an individual’s sex. Accordingly, research illustrated that males and females diverge in the way they process information, shape perceptions (Agosto, 2004), and react to events in the workplace (Konrad, Ritchie, Lieb, & Corrigall, 2000). Gender is defined by the Dictionary (2016) as either the male or female division of a species, especially as differentiated by social and cultural
roles and behavior. According to APA (2012), on the other hand, gender is the attitudes, feelings, and
behaviors that a given culture associates with a person’s biological sex. The latter definition is chosen as it fits this research closely with a more finger-grained focus on the behavior aspect for gender.
Interestingly, literature shows that men and women respond differently when new IT is introduced (Morris, Venkatesh, & Akkerman, 2005; Venkatesh & Morris, 2000; Venkatesh, Morris & Akkerman, 2000). Ahuja and Thatcher (2005) build on it by stating that men and women differ in their propensity to innovate with technology. Accordingly, Barnett and Marshall (1991) indicated in their study that men tend to find their jobs the most salient, while family comes at second place.
21 attributes. Although, Zentner and Mitura (2012) demonstrate that these attributes of males and females have converged as the status of women has risen, much literature still implies the counterpart (Heilman, 2012; Else-Quest et al., 2010; Charles & Bradley, 2009), leaving the following hypothesis behind;
H9d: The effect of transformational IT leadership on ASU will be moderated by gender in such that this effect will be stronger when individuals are female rather than male.
Moreover, as aforementioned, triggers need to be recognized in order for a person to take action (Burke et al., 2006; Langer, 1986). Females are more aware of triggers as they rate the importance of service aspects and physical environment more highly than men (Hofdstede, 1980). Early research (Garai & Scheinfeld, 1968; Parsons & Bales, 1955; William & Best, 1982) also illustrated that females emphasize higher on social cues in forming their behavioral intention to use the IT. However, they also tend to be more anxious than males about system use (Maruping, Likoebe, Magni, & Massimo, 2012; Harris, Jenkins, & Glaser, 2006; Becker & Eagly, 2004; Bozionelos, 1996; Hunt & Bohlin, 1993), which can lead to a lower self-efficacy, possibly lowering perception of ease of use (Venkatesh & Davis, 1996). Low evaluations of ease can cause an increase in the salience in determining user acceptance decisions, giving oneself easier up to employ new features (Eagly & Carli, 1981). Therefore, it is anticipated that for novel situations, females are more salient compared to males. Thus, females more than males could influence the relationship between novel situations and ASU, positively, which gives the following hypothesis;
H9a: The effect of novel situations on ASU will be moderated by gender in such that this effect will be stronger when individuals are female rather than male.
However, because females tend to emphasize higher on social cues, they are also more responsive to social pressures (Harris, Jenkins, & Glaser, 2006; Venkatesh & Morris, 2000), and so to be influenced by others (Becker, 1986; Eagly & Carli, 1981). Males tend to have an overconfident attitude when others evaluate, while women are more accepting of other’s opinions (Venkatesh & Morris, 2000).
22 deliberate initiatives and ASU, the possible moderation effect will be stronger for females rather than males, giving the following two hypotheses:
H9b: The effect of discrepancies on ASU will be moderated by gender in such that this effect will be stronger when individuals are female rather than male.
H9c: The effect of deliberate initiatives on ASU will be moderated by gender in such that this effect will be stronger when individuals are female rather than male.
Age
Despite literature regarding the explanation and prediction process of different genders’ behavior and intentions, the topics for new IT (features) are still underemployed (Malhotra, Galletta, & Kirsch, 2008; Venkatesh & Morris, 2000). Ahuja and Thatcher (2005) detect that it is “an everyday challenge
for managers to find ways of facilitating IT-based innovation and creativity” (Ahuja & Thatcher,
2005; 428) and that age here receives a great deal of attention (Posthuma & Campion, 2009). This limited use of new features for task-related innovation, hinders improving potential IT-related job efficiency, and in turn accomplishing returns for the humongous IT investments (Ahuja & Thatcher, 2005; Hsieh, Rai, & Xu, 2011; Jasperson, Carter, & Zmud, 2005). Therefore, age is taken under the loop to further explain individuals’ perception of IT (Agarwal & Prasad, 1999) and ASU.
The Dictionary (2016) defines a person’s age as a period of human life, measured by years
from birth, usually marked by a certain stage or degree of mental or physical development and involving legal responsibility and capacity. Merriam-Webster (2016) defines age as an individual's
development, measured by years requisite for development of an average person. The first definition is used as it emphasizes the stages of time periods in a person’s life. This research also refers to a certain timeframe (e.g. participants’ average age of 30.18), which this paper investigates.
The influence factor age can come to light when one is asked to use a certain feature or technology – deliberate initiative – to advance him or her task. Czaja and Sharit (1993) explain that users can get more motivated to perform deliberate initiatives if they have a positive attitude towards technology. Furthermore, they explain that certain attitudes belong to particular age groups to which he or she belongs. Accordingly, age can influence the attitudes of a person when asked to use certain (new) system features.
23 their age is one of the most relevant individual factors that influence employees’ perception in the context of system usage (Venkatesh, Morris, Davis, & Davis, 2003). Therefore, it is presumed that younger users, more than older ones, could have a positive effect on the relationship between deliberate initiatives and ASU, leading to the following hypothesis;
H10d: The effect of deliberate initiative on ASU will be moderated by age in such that this effect will be stronger when individuals are of younger age rather than older.
Furthermore, Barnes-Farell and Matthews (2007) illustrate that system usage becomes negative when one has negative thoughts regarding it. This happens for instance, when one experiences the new system or features (novel situations) to be difficult or problematic (discrepancies), happening typically to older employees having less experience with technology as oppose to younger ones (Elias, Smith & Chet, 2010). Therefore, it is anticipated that younger employees, more than older ones, could influence the relationship between discrepancies and ASU, building the following hypothesis:
H10b: The effect of novel situations on ASU will be moderated by age in such that this effect will be stronger when individuals are of younger age rather than older.
This same reasoning is in line for discrepancies, where technology that is perceived as unwanted or not obligatory, can cause a feeling of overload (Sun, 2012), which tends to happen more to older, than younger users due to their level of experience with systems (Elias, Smith & Chet, 2010). Thus, it is presumed that younger employees, more than older ones, could influence the relationship between discrepancies and ASU, positively, giving the following hypotheses;
H10c: The effect of discrepancies on ASU will be moderated by age in such that this effect will be stronger when individuals are of younger age rather than older.
24 anticipated that for the relationship between transformational IT leadership and ASU, the possible moderation effect will be stronger for individuals who are of older age, rather than young, allowing the following final hypothesis to be stated;
H10a: The effect of transformational IT leadership on ASU will be moderated by age in such that this effect will be stronger when individuals are of older age rather than young.
3. Methodology
The methodological part starts with the sample and date collection, where the collection of data and preparation is revealed. The analysis of this data is explained in the second paragraph.
3.1 Sample and Data Collection
Data for this research is gathered from IT employees. The participants were asked to fill in the survey, which with the help of survey instruments was directed towards personal innovativeness in IT, age, gender, facilitating conditions, adaptive systems use (trying new features, feature substituting, feature combining, and feature repurposing), and the triggers (novel situations, other’s people use, discrepancies, deliberate initiatives, and transformational IT leadership).
Pretest
When the questionnaire was finished, a double check was held with another senior researcher to search for errors. In addition, to test the time, a pretest took place, which took around ten minutes. The estimation of the survey was set from five to twenty minutes. This eliminated some participants finishing within two minutes. Another reason why some candidates where disregarded was the inconsistency regarding some questions, as some questions were written opposed and candidates filled in inconsistent answers, like the extreme opposite on those ends (e.g. pushed in the last bullet point on the same question written opposed later on).
Participants
25 Table 3
Demographic characteristics of the Sample
Variable Sample composition Team members (n=120) in %
Age Gender
Mean = 30.18; Std. dev. = 6.657; range 21-54 years
Female 45 37.5%
Male 75 62.5%
Language Dutch 25 20.8%
English 95 79.2%
Team size Mean = 8.17; Std. dev. = 8.590; range 0-55 people per team
Tenure Mean = 2.88; Std. dev. = 2.030; range 0-10 years
Participants where gathered to take the survey, either in person, on social media, or via mail. Many organizations where approached by gathering their contact information from the electronic database Orbis. The majority of the data however is gathered through connections via social media (LinkedIn and Facebook), friends and family members that are working or have a business of their own, and places that the researchers had previously worked or followed an internship. Of the 5298 organizations, 167 agreed to participate, indicating a response rate of 3.2 percent. Of those respondents, 47 were deleted due to the elimination methods described above. Accordingly, the final sample consisted of 120 valid responses, as these surveys were completed fully and within a reasonable amount of time, making the actual response rate 2.3 percent.
45 of these are female participants, while 75 are male. The average age was 30,18 years, which fluctuated between 21 being the lowest, and age 54 being the highest. Furthermore, the highest obtained education level of the participants was questioned. The highest percentage of respondents are from the Bachelor’s degree (38%), followed by respondent with a Master’s degree (35%), while 16% possesses a professional degree, and 8% managed to get a high school diploma. Lastly, a minority (3%) filled in the option: ‘other’. These percentages are displayed in figure 2.
Figure 2
Highest Education Level Obtained
26 Moreover, firms in the final sample operate in energy services (1%), financial and/or insurance services (17%), health care services (4%), legal services (1%), production (9%) and public services (1%). Lastly, 67 percent of the participant indicated to work in a different industry then provided in the questionnaire. Most of them wrote down that they are working in the IT industry.
Figure 3
Industry Sector
Reasoning of Definitions
Definitions that are in the original model of Sun are captivated directly from his article. This is firstly because barely any other literature defines items in Sun’s research model differently. Secondly, Sun’s (2012) model is take ‘as is’ which asks for the hypothesis to remain unchanged, and so the definitions, as otherwise the consistency and conclusions of re-testing and comparing the original hypothesis to the ones used in this research would be very low. By keeping it ‘as is’, the research can compare the data from Sun (2012) to draw conclusions as they have the same meaning. This is thirdly, because removing insignificant hypothesis from the model could make something significant, insignificant, or vice versa, also because certain triggers only work in combination of one another. Forth, it is kept as the questionnaires are taken directly from multiple authors, for which changes in definition could disrupt the consistency in questions, which could lead to the same (in)significant story mentioned in point three. Lastly, the definitions define the research purpose of this specific subject most accurately since the modelling is build with this line of thinking.
17% 1% 4% 1% 9% 67% 1%
Financial and/or insurance services
Legal services Health care services Energy services Production Other
27 Measures
To come off with reliable data, measures of existing questionnaires, of previously peer-reviewed articles are used. The survey existed of twelve open questions, including demographic questions, and closed questions consisting of a seven-point scale (completely disagree to completely agree). This choice was adopted from Lozano, Garcia-Cueto and Muniz (2008), as they concluded that a broader set of items increases both the reliability as the validity of the data. One open question asked for a specific IT situation that the participant experienced in order the answer the upcoming questions. Appendix B includes each survey instrument and the authors from which each question is gathered.
Dependent variable
ASU is only possible with one or a combination of the triggers aforementioned, making ASU the main dependent variable (Sun, 2012) of this research. This will be measured with the construct Sun (2012) developed for this variable specifically. One of these four triggers, discrepancies, also takes the role of a dependent variable, as its’ existence depends on novel situations and deliberate initiative.
Independent variables
The other three triggers are the independent variables, as indicated in Table 2. Where Sun (2012) indicated a specific IT tool in his questionnaire to be described, this questionnaire asked for any IT tool in general that the researcher has used. Furthermore, the independent variable transformational IT leadership is added to analyze if it would relate as a trigger for ASU. This is added based on the revised questionnaire regarding transformational leadership of (Podsakoff et al., 1996), which Biernath (2014) and Sietsma (2014) revised by directing it to IT specifically. For example, instead of “My team
leader fosters collaboration among work groups“, Biernath (2014) and Sietsma (2014) formulated
“My team leader fosters collaboration between teams by using IT tools”.
Moderating variables
Between the dependent and independent variables of ASU, are the moderators: personal innovativeness in IT, facilitating conditions, age and gender placed. The latter two are added by the researcher to see whether these demographics serve have a moderation effect between the triggers and ASU. Each of the moderating variables consists of four links to the four triggers, indicating a total of sixteen moderation effects. The first moderator personal innovativeness in IT is taken from Agarwal and Karahanna (2000), whereas the second one, facilitating conditions, is subtracted from Venkatesh, Morris, Davis and Davis (2003).
Control variables
28 variable comes is taken from Schaubroek and Chan (2007), and Wang, Li, and Hsieh (2013). The latter author also indicated control variables for tenure, age and education. The range for education is measured to receive a more accurate view of range of respondents.
3.2 Analysis
The analysis starts with the explanatory factor analysis (EFA) which was made in SPSS Statistics 23 using an explanatory factor analysis with varimax rotation. The main purpose for this was to gather information regarding the relationship between different items of the constructs and their scores (Hinton, 2004). If an item contained a loading of at least 0.5, which is the criteria of Song, Bij and Song (2011) to be included, they remained. In table 4 the results are displayed, which show that not all items of the constructs are included (TLAV 2, for example, is disregarded). Based on these results, an EFA was separately conducted for each construct named in measures. Another removal took place for items not loading or loading into more than one construct, leading to the results displayed in table 4.
Table 4
Explanatory factor analysis loadings and Cronbach’s alpha (Transformational IT Leadership)
Construct Item Factor Loading Cronbach’s alpha
Transformational IT Leadership
Articulating an IT-vision TLAV 1 .802 α = .803
TLAV 3 .744
Providing an appropriate IT-role model TLAM 2 .847 α = .919
TLAM 3 .785
Fostering the acceptance of group goals through IT TLFG 1 .863 α = .815
TLFG 2 .779
High performance expectations with IT TLPE 2 .944 not applicable
Individualized support TLIN 2 .918 α = .860
TLIN 3 .870
Intellectual stimulation with IT TLIS 2 .991 not applicable
Similarly, the remaining variables are also gathered in the EFA for ASU (separately), and for transformational IT leadership combined with novel situations, discrepancies (triggers), personal innovativeness in IT, and facilitating conditions (moderators) all together. This analysis was followed by a confirmatory factor analysis (CFA) in Lisrel 8.80 for which the results are displayed in table 5.
29 Table 5
Confirmatory factor analysis loadings, t-values, and Cronbach’s alpha
Construct Item Factor Loading t-value Cronbach’s alpha
Adaptive System Use
Trying new features TR1 1.11 10.21 α =0.835
TR2 1.05 8.44 TR3 1.20 11.61 TR4 0.76 7.12 Feature substituting FS1 1.45 11.90 α =0.875 FS2 1.33 10.27 FS3 1.24 10.61 Feature combining FC2 1.71 13.15 α =0.888 FC3 1.43 10.99 FC4 1.36 9.85 Feature repurposing FR2 1.67 12.98 α =0.917 FR3 1.49 11.99 FR4 1.54 12.38 FR6 1.47 9.50
χ ² = 129.97; df. = 71 ; RMSEA = .084; NFI = .94; CFI = .97; GFI = .87 Novel situations
Changes in System Environments SE1 1.35 5.39 α = 0.846
SE2 2.71 7.24
Other People’s Use OU2 1.46 9.67 α = 0.864
OU3 1.04 7.01
Discrepancies DP1 1.64 10.76 α = 0.905
Transformational IT Leadership
DP2 1.76 11.48
Articulating an IT-vision TLAV1 1.15 10.86 α = 0.803
TLAV3 0.90 10.01
Providing an appropriate IT-role model TLAM2 1.19 12.29 α = 0.919
TLAM3 1.37 13.62
Fostering the acceptance of group goals through IT TLFG1 1.09 10.74 α = 0.815
TLFG2 0.88 9.81
Individualized support TLIN2 0.84 9.10 α = 0.860
TLIN3 0.95 11.03
Personal Innovativeness in IT PIIT1 1.22 10.55 α = 0.859
PIIT3 1.37 10.27
PIIT4 1.03 10.15
Facilitating Conditions FCOND1 0.80 5.81 α = 0.743
FCOND3 1.04 6.61
χ ² = 128.77; df. = 116; RMSEA = .030; NFI = .93; CFI = .99; GFI = .90 RMSEA - Root Mean Square Error of Approximation; NFI - Normed Fit Index; CFI - Comparative Fit Index; GFI - Goodness of Fit Index
For ASU it illustrated: χ2 = 129.97, df = 71 Root Mean Square Error of Approximation (RMSEA) = .084, Normed Fit Index (NFI) = .94, Comparitive Fit Index (CFI) =.97, and Goodness of Fit Index (GFI) = .87. The scores for the next goodness of fit statistics are: χ2 = 128.77, df. = 116, RMSEA = .030, NFI = .93, CFI = .99, and GFI = .90, existing of transformational IT leadership, together with the independent and moderating variables.
