Theory, Research Questions, Conceptual Model and Hypothesis

End User Computing Support

To measure the use of different sources of support, EUC needed a more precise definition. Many studies show different perspectives on EUC support (Arnoudse & Oulette, 1986; Bruton, 1995; Doll & Torkzadeh, 1993; Heie & Heistad, 1998; Larsen, 1989; Smith, 1997; Winter, Chudoba & Gutek, 1997).

Through a thorough analysis of the different perspectives on EUC support, a partitioning of EUC support was needed. Doll and Torkzadeh (1993) divides EUC support into three categories. These are:

• Consultation

• Training

• Documentation

This survey seeks to measure ad hoc support needs. The category Training is therefore irrelevant. Consultation and Documentation were singled out as the types of EUC support that would be tested for in this survey. Further analysis showed that Consultation and Documentation could be divided in formal vs. informal sources of support and personnel vs. impersonal sources of support. This resulted in four different types of EUC support sources:

• Personal and informal consultation with colleagues.

• Personal and formal consultation with computer experts.

• Use of external documentation (impersonal and informal).

• Use of internal documentation (impersonal and formal).

Through this review EUC support was defined to be:

All sorts of IT-help that an end user receives or uses in his work to solve arising problems or acquire expertise and skills within IS-use, so that they easier can achieve organizational goals.

This definition limits the perspective on EUC support and makes it somewhat easier to measure.

End User Qualities

As the purpose of this study is to find out whether different end user qualities can explain the differences in their choice of support sources, it is equally important to find these qualities.

There exists some literature on EUC support, but not very much on the end user's choices of support depending on his basic qualities (i.e., skills, etc.). Winter et al. (1997) concluded in their survey that even though training and support could have improved the end user's computer knowledge, it is clear that it has not lead to high computer knowledge. Their opinion is that it is important for the support personnel to have some knowledge about the end user's computer skills to give them proper support. It then seems reasonably obvious that computer skills might affect the end user's choice of different support services. I therefore ask:

Do IT-skills influence the end user's choice of support services?

Table 1: EUC Support Categorization

EUC support	Informal	Formal
Personal	Consultation with colleagues, or other non professional IT workers	Consultation with IS-professionals
Impersonal	Use of external documentation not developed by the local IC. This could be manuals, periodicals, etc.	Use of internal documentation developed by the local IC

One would believe that end users with low computer knowledge and skills would need more support than those with high computer knowledge and skills. Øystein Sørebø wrote a paper in 1996 called: "End-User Computing and the perceived need for support services: Toward an explanation of the independent-user paradox." The qualities he believed to affect the perceived need for support services include: IT-involvement, computer self-efficacy, and informational influence (from colleagues).

Sørebø questions whether the end user's IT-involvement might have a significant influence on the perceived need for support services. Earlier studies have shown that involvement affects information searching (Laurent & Kapferer, 1985; Zaichkowsky, 1986). Finding the solution to computer related problems, through the use of different support sources, could easily be compared with information searching. Zaichkowsky (1986) also points out that an individual's attention towards and experience of what's important in relation to the execution of a specific behaviour will vary with the individual's involvement. In this context, execution of a specific behaviour can be compared with the use of different sources of support and the individual's involvement could be different aspects of the end user's involvement toward the computer.

On these basis one could ask:

Do IT-involvement influence the end user's choice of support services?

Computer Self-Efficacy is an important end user quality. Compaeu and Higgins (1995) argues that this special psychological state will affect the end user's belief about his need for support services. Belief about the need for support services and actual use of different support services are clearly related topics, and therefore my question is:

Do computer self-efficacy influence the end user's choice of support services?

Now I will turn to a more detailed description of each of the three explanatory factors.

IT-Skills

The concept IT-skills is not easily defined. IT is widely used, but often without providing a precise definition. Much work is done on the related concept End User Computing Sophistication. The reason why I have not used the concept, End User Computer Sophistication, is that different authors have defined it differently in different surveys (Blili, Raymond & Rivard, 1994; Huff, Malcolm & Marcolin, 1992; Marcolin, Munro & Compeau, 1993; Rockart & Flannery, 1983; Zinatelli, 1996). It would be difficult to compare the results from the different surveys because of the variations in the definition of the concept.

The subject skill is often connected to the subject's ability. A few researchers (Cheney & Nelson, 1988; Koohang et al., 1992; Marcolin et al., 1996) have tested end user ability. Both Marcolin (1996) and Koohang (199x) have used Cheney and Nelson's instrument for developing their instruments on end user abilities. Cheney and Nelson identified three clear factors within end user computing abilities: technical abilities, modelling abilities and application abilities. Technical abilities apply to programming, the use of hardware and operating systems. Modelling abilities apply to subjects regarding software engineering. Application abilities apply to skills that are most typically associated with the use of applications systems. All these factors are important for measuring end-users' IT-skills. The aim of this study was, however to measure work-relevant IT-skills. The measure of technical and modelling abilities was therefore less interesting. On this basis, I defined IT-skills to be:

In what degree a person manages to solve different problems with help from different work-relevant information system tools.

IT-Involvement

Earlier research on IT-involvement has mostly been about participator behaviour within IS-development (Ives & Olsen, 1994). The psychological dimension of this participation has been brought to focus in the later years. In spite of Barki and Hartwick (1989), Kappelman (1990) and Kappelman and McLean (1993, 1994) trying to establish a conceptual partitioning between participation and engagement as two aspects of involvement, it is still common to use end user involvement as a description of participant behaviour (Doll & Torkzadeh, 1994; Igbara & Guimaraes, 1994). A solution to this partitioning of behavioural and psychological involvement is to denote them both end user involvement, and to distinguish between the two components situational involvement and intrinsic involvement (Jackson et al., 1997). One can further divide intrinsic involvement in a psychological condition and as involvement towards information technology, the computer and software or involvement towards a process. My aim with IT-involvement is to measure involvement towards information technology, the computer and software. Table 2 shows the partitioning of the concept.

Table 2: End User Involvement Partitioning

End User Involvement	Related to the Phenomenon	Can be Divided Into:
Situational Involvement (End User Participation)	Behaviour	Process Participation or System Usage
Intrinsic Involvement (End User Engagement)	Psychological State	Involvement Towards Information Technology, the Computer and Software or Involvement Towards a Process

With basis in the work of Barki and Hartwick (1989), I have defined IT-Involvement as follows:

The importance and personal relevancy an end user attaches to a computer and the use of it.

Computer Self-Efficacy

Compeau and Higgins (1995) did a survey on the concept of self-efficacy to prove its usability in the attempt to understand individual behaviour towards computers. The term self-efficacy is future-oriented. It does not deal with what a person has done earlier, but rather with a person's beliefs of what can be done in the future (Compeau & Higgins, 1995b, p. 192).

It is "borrowed" from social psychology, where self-efficacy is said to be the user's beliefs about his capability to organize and execute the courses of action required to manage prospective situations (Bandura, 1996).

Self-efficacy has its origin in the writings of Albert Bandura (1986, 1995). He defines it to deal with: "peoples judgement of their own capabilities to organize and execute courses of action required to attain designated types of performance. It is concern not with the skills one has, but with the judgements of what one can do with whatever skills one possesses" (Bandura 1986, p. 391). Thus, Computer Self-Efficacy represents an individual's perception of his ability to use computers in the accomplishment of a task (Compeau & Higgins, 1995a).

The concept has three dimensions (Compaeu & Higgins, 1995a, 1995b). These dimensions are: magnitude - the level of computing task difficulty the user can attain; strength - whether the conviction regarding magnitude is strong or weak and generalizability - the degree to which the expectation is generalized across different software packages and different computer systems.

End users with a high magnitude of Computer Self-Efficacy might judge themselves as capable of operating with less support and assistance than those with lower magnitude of self-efficacy (Compaeu & Higgins, 1995a, 1995b).

Compeau and Higgins (1995b, p. 195) show that support was negatively related to self-efficacy with a regression coefficient of -0,16. The survey thereby showed that the more support given to the end user the less computer self-efficacy he possessed.

Following these research questions, conceptual definitions and discussions, I will utilize the model in Figure 1.

Figure 1: Research Model

Hypothesis:

H1:	The end user's IT-skills will covariate with their respective source of support choices.
H1a:	High IT-skills is negatively related to the use of formal sources of support.
H1b:	High IT-skills is positively related to the use of informal sources of support.
H1c:	High IT-skills is negatively related to the use of internal documentation.
H1d:	High IT-skills is positively related to the use of external documentation.
H2:	The end user's Computer Self-Efficacy will covariate with their respective source of support choices.
H2a:	A high degree of Computer Self-Efficacy is negatively related to the use of formal sources of support.
H2b:	A high degree of Computer Self-Efficacy is negatively related to the use of informal sources of support.
H2c:	A high degree of Computer Self-Efficacy is negatively related to the use of internal documentation.
H2d:	A high degree of Computer Self-Efficacy is positively related to the use of internal documentation.
H3:	The end user's IT-involvement will covariate with their respective source of support choices.
H3a:	High IT-involvement is positively related to the use of formal sources of support.
H3b:	High IT-Involvement is positively related to the use of informal sources of support.
H3c:	High IT-Involvement is positively related to the use of internal documentation.
H3d:	High IT-Involvement is positively related to the use of external documentation.