An Empirical Examination of the Impact Organizational ... - CiteSeerX

Employees’ Computer Self-Efficacy 1

Chapter I

An Empirical Examination of the Impact Organizational Culture Has on Employees’ Computer Self-Efficacy Yihua Sheng, Southern Illinois University, USA J. Michael Pearson, Southern Illinois University, USA Leon Crosby, Grand Valley State University, USA

ABSTRACT IT-based business initiatives, such as ERP and BPR, require high computer self-efficacy among employees because as changes require large-scale use of computers. Computer self-efficacy is affected by many internal and external factors; for instance, personality or organizational culture. While extensive literature exists on how psychological and sociological factors affect a person’s self-efficacy, almost no research has been done on how organizational culture could influence employees’ computer self-efficacy. This chapter examines the relationship between organizational culture and employees’ self-efficacy for a sample of 352 subjects. The results, from multiple regression and discriminant analysis, show teamwork and information flow contribute most to employees’ computer self-efficacy. Copyright © 2004, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.

2 Sheng, Pearson & Crosby

INTRODUCTION Over the past two decades, computer usage has increased dramatically in business operations. Applications such as Enterprise Resource Planning (ERP), Business Process Reengineering (BPR), Supply Chain Management (SCM) and Customer Relationship Management (CRM) require an extensive use of computer technology. Few of these applications were completely successful and few were total failures, with the rest falling somewhere in-between (Kotter, 1995). The technology has been proven by many successful implementations; however, researchers want to know how the overall success rate could be higher. They have explored characteristics such as employees’ computer self-efficacy, organizational culture and structure, management style and readiness of an organization (Al-Khalifa and Aspinwall, 2000; Cabrera, Cabrera, and Barajas, 2001; Hoffman and Klepper, 2000; Kim, Pindur, and Reynolds, 1995; McNabb and Sepic, 1995; Stock and McDermott, 2000). Computer self-efficacy refers to one’s belief in one’s ability to apply his or her computer skills to a wide range of tasks (Compeau and Higgins, 1995). There is a consensus among researchers and practitioners that computer self-efficacy is positively related to an individual’s attitude towards information technology. A detailed list of empirical studies incorporating self-efficacy in the conceptual and/or research models can be seen in Agarwal, Sambamurthy and Stair (2000). Computer self-efficacy has been found to be positively related to performance in software training (Gist, Schwoerer, and Rosen, 1989), perceived ease of use of computer systems (Agarwal et al., 2000; Hong, Thong, Wong, and Tam, 2002; Hung, 2003; Igbaria and Iivari, 1995; Venkatesh, 2000; Venkaresh and Davis, 1996) and adaptability to new computer technology (Burkhardt and Brass, 1990). All of these, in turn, influence the successful deployment of an information system. Several studies have examined the relationship between organization characteristics and employees’ behavior; for instance, the relationship between organizational climate, which is a manifestation of culture (Schein, 1985) and employee involvement (Shadur, Kienzle, and Rodwell, 1999; Tesluk, Vance, and Mathieu, 1999). Studies, most of which are psychological and sociological in nature, have been conducted to identify the determinants and antecedents of self-efficacy and computer self-efficacy (Bandura, 1977; Compeau and Higgins, 1995; Gist et al., 1989; Thatcher and Perrewé, 2002). Unfortunately, very little research has been done on the macro level to see how organizational culture could affect and shape employees’ computer self-efficacy. This study empirically investigates the relationship between organization culture and an employee’s computer self-efficacy.

Copyright © 2004, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.


ORGANIZATIONAL CULTURE Environmental influences such as social pressure and personal factors such as personality and behavior are reciprocally determined. This has been identified as a “triadic reciprocally” by Bandura (1977, 1982). Organizational culture is one type of environmental influence which impacts the way people (employees) think, perform tasks and communicate/interact with each other. According to Schein (1985), culture is “a pattern of basic assumptions — invented, discovered, or developed by a given group as it learns to cope with its problems of external adaptation and internal integration — that have worked well enough to be considered valid and, therefore, to be taught to new members as the correct way to perceive, think and feel in relation to those problems” (p. 9). Also, Hofstede (1984) describes organizational culture as “the way things are done in the business.” More specifically, organizational culture is the “shared perceptions, patterns of belief, symbols, rites and rituals and myths that evolve over time and function as the glue that holds the organization together” (Zamanou and Glaser, 1994). Based on these definitions, it is easy to see that the existing culture of an organization provides a corporate framework that provides guidance on issues like how work is done, the use of technology, how people think and standards for interaction and communication. The shared perceptions and beliefs that make up an organization’s culture are fostered and cultivated by communications and interactions among people inside and outside the organization. These perceptions and beliefs then effect and can be influenced by people’s behaviors on things like how to solve problems, how to conduct a job and how to communicate (Bates, Amundson, Schroeder, and Morris, 1995). These, in turn, affect an individual’s job performance and satisfaction, and then affect a firm’s performance. It has been shown that organizational culture (and various subcultures within the organization) can have a positive effect on competitive advantage, increased productivity and a firm’s performance (Yeung, Brockbank, and Ulrich, 1991). On an individual’s level, Zamanou and Glaser (1994) found organizational culture could affect an employee’s participation and involvement. Work is done by people who make up an organization, not by the organization itself. Organizational culture is ultimately manifested, represented and maintained by sense-making efforts and actions of individuals (Harris, 1994). If organizational culture impacts a firm’s performance or productivity, it is because the organizational culture impacted individuals first, which in turn affected a firm’s overall performance, productivity or competitive advantage. Several studies have shown how organizational culture or changes in organizational culture can facilitate or hinder business change initiatives such as BRP, ERP and TQM (Al-Khalifa and Aspinwall, 2000; Bennett, Fadil, and Greenwood, 1994;



Detert and Schroeder, 2000; Hoffman and Klepper, 2000; Kim et al., 1995; Wayne, Mooney, and Sheldon, 1999). We, in this study, will investigate how, on a macro level, an organization’s culture affects their employees’ behavior when using computers, as proxied by an individual’s computer self-efficacy.

COMPUTER SELF-EFFICACY Many individuals believe that implementing significant initiatives like TQM, ERP or BPR involves a change in organizational culture (Kim et al., 1995; McNabb and Sepic, 1995). For example, the introduction of large-scale computer technologies in business operations can drastically change the way a company conducts its business. Many jobs, processes, procedures, communications and evaluation criteria must be redefined (Cabrera et al., 2001). Many tasks will be performed on technologies using software packages brought in by ERP, BPR and CRM systems. During the implementation of these applications, employees may need to switch from paper-based work to computer-based work, or change from one kind of software application to another new or different application. Under these circumstances, an employees’ computer self-efficacy would be an important issue that needs to be considered before their company decides to implement these information systems. Questions like, “Are our employees ready to use computers and software packages in their daily work?” and “Are our employees confident in their ability to use these computers and software packages in their daily work?” need to be evaluated and answered before rolling out the project. Together with research regarding the consequences of computer selfefficacy, there is a related body of literature concerning the determinants and antecedents of computer self-efficacy. Past performance and the degree of system use have been found to have a significant impact on a person’s computer self-efficacy (Henry and Stone, 1999). Compeau and Higgins (1995) identified encouragement by others, others use of computers, and organizational support as factors having a significant influence on employees’ computer self-efficacy. Gardner and Rozell (2000) provided a comprehensive list of determinants of a person’s computer self-efficacy based on Gist and Mitchell’s (1992) work. Gardner and Rozell classified the determinants of computer self-efficacy into quartiles according to the variability (low/high) and locus (external/internal) of these determinants. Factors including ability and personality are internal with low variability; while factors such as watching others (modeling), persuasion and feedback have an external locus and high variability. Computer self-efficacy represents a comprehensive judgment of one’s ability to perform a task. It is not a static or stable trait; rather a situation-specific, dynamic judgment that changes with acquired information, such as the change of environmental settings or the change of task conditions and feedback (Gist and Mitchell, 1992). Stable Copyright © 2004, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.


personal traits, such as negative affectivity and computer anxiety, and stable situation-specific individual differences such as personal innovativeness, have been found to be the antecedents of computer self-efficacy (Thatcher and Perrewé, 2002). While changing internal determinants could be difficult, companies seeking to improve an employee’s computer self-efficacy could take steps to influence the external determinants. For example, an organization could enhance the spirit of teamwork to promote a modeling effect, or increase information flow to improve performance feedback. Many of these external factors are part of, or can be influenced by, organizational culture. IT-based changes such as ERP, BPR and CRM can change an organization’s culture along with large scale usage of computer technology and/or applications in an employees’ daily work; thus, to improve an employees’ computer self-efficacy, changing and/or modifying and enhancing an organization’s culture, could increase the chance of success of these IT-based initiatives.

RESEARCH MODEL AND HYPOTHESES Our research model is presented in Figure 1. Of the many measures available, we used the Organizational Culture Scale (OCS) as developed by Glaser and Zamanou (1987). Their scale consists of six components grounded in both management and communication research: teamwork, climate and morale, supervision, involvement, information flow and meetings. Initially, these items were developed from a review of literature and interviews with employees at a Northwestern wood products company. After revision, the instrument was shown to be a valid and reliable measure of the intended constructs. The authors of this study do not claim that these six sub-constructs are a comprehensive set of measures for organizational culture, but that these do represent important elements in the overall makeup of an organization’s culture. The computer self-efficacy construct of an individual was assessed by using the 10-item scale from Compeau and Higgins (1995). Compeau and Higgins (1995) developed this scale based on Social Cognitive Theory and other existing scales. It is a task-focused, rather than skill-focused, measure. The reliability of the scale was reported to be .95 in their study. Within the OCS, teamwork was defined as “reported coordination of effort, interpersonal cooperation and/or antagonism, resentment, power struggles within sections or divisions; people talk directly and candidly about problems they have with each other” (Glaser and Zamanou, 1987). Teamwork encourages cooperation and coordination. Within a positive teamwork environment, members talk directly and work together; the close working relationship facilitates the modeling and persuasion effect of computer self-efficacy. In addition, getting encouragement and feedback from others is natural within a cooperative and coordiCopyright © 2004, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.


Figure 1. The Research Model Teamwork Climate and Morale Supervision

Computer Self-efficacy

Information flow Involvement Meeting

nated atmosphere. Within a supportive teamwork environment, members are willing to help each other; they do not feel isolated when performing tasks on computers. They can learn computer skills from each other and receive training conveniently. This would improve their computer self-efficacy (Locke, Frederick, Lee, and Bobko, 1984). Additionally, since this type of informal training is more of a behavioral modeling style than tutorial style, training could contribute more to the team member’s computer self-efficacy (Gist et al., 1989). Thus, we propose the following: Hypothesis 1: Teamwork is positively related to an employees’ computer selfefficacy. Climate and morale is defined as “reported feelings about work conditions, motivation, general atmosphere, organizational character” (Glaser and Zamanou, 1987). People having higher morale are usually more self-motivated to overcome obstacles (Gardner and Rozell, 2000; Gist and Mitchell, 1992). When they face difficulties in the workplace, the motivated employees usually try harder to overcome these obstacles by actively seeking supporting resources such as advice or training. The training and the resultant job performance achieved would contribute to his or her self-efficacy (Bandura, 1977; Henry and Stone, 1999; Locke et al., 1984). Therefore, the following hypothesis is proposed: Hypothesis 2: Organizational climate and employees’ morale are positively related to employees’ computer self-efficacy. Supervision is defined as “reported information by the employees on their immediate supervisor; the extent to which they are given positive and negative feedback on work performance; the extent to which job expectations are clear” Copyright © 2004, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.


(Glaser and Zamanou, 1987). The evaluation of a supervisor is one kind of feedback. Efficacy beliefs and performance are more likely to rise with feedback than without, no matter if it is positive or negative (Karl, O’LearyKekky, and Martocchio, 1993). Positive supervision could boost people’s confidence, encourage employees to use the computer even if they have only limited experience with the computer and would encourage them to be willing to switch to computer-based work or a new kind of software package. This would improve the employees’ computer self-efficacy as past performance can contribute to one’s self-efficacy (Banduara, 1977; Henry and Stone, 1999; Locke et al., 1984). These arguments suggest the following: Hypothesis 3: Supervision is positively related to an employees’ computer selfefficacy. Information flow is defined as “links, channels, contact, flow of communication to pertinent people or groups in the organization; feelings of isolation or being out of touch” (Glaser and Zamanou, 1987). Good information flow could help propagate information concerning the use of computers by peer workers, how the computer or new software package was used by others, what kind of help or benefit the computer or new software package provided, how much other people benefited from using the computer or new software package, etc. By supporting good information flow, the use of computers and new ways of using a computer would promote an employee’s computer self-efficacy. Good information flow also promotes a feedback system. Employees could use feedback as a resource for performance monitoring and adjust their behavior accordingly to achieve better outcomes that in turn would improve their computer self-efficacy. As mentioned previously, computer self-efficacy is a dynamic judgment that changes with newly acquired information. Employees in an environment with good information flow could construct and orchestrate adaptive performances to meet changing situational demands. The adoption of new beliefs and their capability to perform a task could be facilitated by the availability of accurate information about the causes of performance, as well as information about the specific tasks that the employee is undertaking (Gist and Mitchell, 1992). This suggests the following hypothesis: Hypothesis 4: Information flow is positively related to an employees’ computer self-efficacy. Involvement is defined as “reported input and participation in decision making; respondents feel that their thoughts and ideas count and are encouraged by top management to offer opinions and suggestion” (Glaser and Zamanou, 1987). People involved in decision-making processes are usually well informed. Copyright © 2004, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.


During decision-making, participants would become aware of the pros and cons of the computer technology and the software application. This, to some extent, enhances information flow. The suggestions, presentations and debates are part of the education, persuasion and feedback process for people involved in the decision-making processes; these could all positively contribute to their computer self-efficacy. In addition, involvement is another kind of affirmation of the employees’ achievement from their supervisors. Usually, only people with good/positive contributions are involved in the decision making process. Supervisors tend to listen to people who are knowledgeable in their fields, do things right and have superior past performance. These people’s ideas and thoughts, such as the usefulness of a particular computer technology or software application, the benefit of using a computer technology or application, etc., are usually solicited, assessed and possibly adopted by supervisors. Based on these typical scenarios, involvement in the decision-making process can be viewed as a positive feedback from the supervisors, which would improve these employees’ computer self-efficacy. Thus, the following hypothesis is proposed: Hypothesis 5: Involvement is positively related to an employees’ computer self-efficacy. Meetings are defined as “reported information on whether meetings occur and how productive they are” (Glaser and Zamanou, 1987). Periodical meetings can be regarded as a way to enhance information flow and involvement, which all impact self-efficacy. Various opinions and suggestions from employees and management are discussed in meetings. People can obtain a variety of information, either educational or noneducational and feelings of involvement and teamwork through meetings, which could in turn influence people’s selfefficacy. This leads to the following hypothesis: Hypothesis 6: The occurrence and productivity of meetings is positively related to an employees’ computer self-efficacy. Although teamwork, climate and morale, supervision, information flow, involvement and meetings all could positively impact an employees’ computer self-efficacy, the degree and direction of the impact from each of the these subconstructs could be different. Employees might react more to some subconstructs than the others might. Thus, we proposed the following hypothesis: Hypothesis 7: Some sub-constructs of organizational culture impact an employees’ computer self-efficacy more than the others.



RESEARCH METHODOLOGY Subjects The target population for this study was knowledge workers — specifically, individuals who create information and knowledge as part of their daily work activities. Knowledge workers, unlike clerks, use a variety of software, such as database software, analytic software, word processing, etc. to complete their tasks (Turban, McLean, and Wetherbe, 2002). The method of using the software is not structured and/or predetermined. They have discretion as to what and how to use this software to complete their tasks. This implies that knowledge workers have the willingness and capability to discover, learn and master the functionalities of the software they use. Therefore, their level of computer self-efficacy is an important variable in their success. Representatives from 20 companies were identified and asked to participate in this study. The companies were large, multinational organizations that represented a diverse group of industries including agriculture, insurance, oil refining, consulting, transportation and finance. Each representative was asked to distribute 20 questionnaires to a randomly selected group of knowledge workers throughout their organization. Individuals were identified for participation based on their job description and a short interview to determine the extent to which they utilized computer technology in their daily activities.

Instrument The survey package contained a cover letter from the organization’s representative, a letter from the researchers explaining the purpose of the study and the questionnaire. All respondents were guaranteed confidentiality of their responses. This study used three sections of a multipart questionnaire - ten questions were included to solicit information about the respondent and their organization; 31 questions designed to measure the six constructs associated with organization culture; and finally, 10 questions to measure the respondent’s level of computer efficacy. As a follow-up after two weeks, the company representatives contacted those individuals who had not completed the survey instrument. Three hundred and fifty two individuals completed the survey instrument for a response rate of 88 percent (352/400).

RESULTS Three hundred and fifty two subjects voluntarily participated in this study. Gender was split equally among male and female respondents. Fifty-nine percent were college graduates with fifty-four percent citing business as their primary educational background. Seventy-one percent of the respondents were employed in a functional area other than information systems. Sixty-eight percent Copyright © 2004, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.


Table 1. Summary of Key Demographics Functional Area: Information Systems Accounting

28.70% 21.00%

Marketing and Sales Human Resources

13.10% 6.00%

Management Other Areas

6.00% 25.20%

Gender: Male Female

50.00% 50.00%

Educational Level: Completed High School Some College

12.50% 10.50%

College Degree Position in the Organization: Executive

3.10%

Middle Management First Line Management Professional Technical Clerical

17.90% 21.60% 31.30% 12.80% 9.10%

Other

58.80%

Some Graduate Work Graduate Degree

8.50% 9.70%

Educational Background: Business Other

54.00% 46.00%

7.40%

of the participants were either middle management, first line management or professionals. Almost all (99 percent) responded that they used a computer at their place of work several times a day. Obviously, the use of computers is an integral part of their jobs. In fact, 92 percent of the respondents reported that the use of a computer was required at their jobs. Therefore, the participants of this study were predominantly knowledge workers. A summary of the key demographic characteristics is presented in Table 1.

Analysis Before we tested the hypotheses set forth in this study, we ran a confirmatory factor analysis (principal components, varimax rotation) on the organizational culture construct. The rotated factor matrix indicated that all the items loaded as expected with good convergent and discriminant validity. Construct reliability or internal consistency was assessed using Cronbach’s alpha. These values ranged .85 to .94, thus indicating strong construct reliability (see Table 2). Table 3 shows the reliability of and correlations among the tested variables. The reliabilities are all .78 or above. Hypotheses 1 through 6 are examined using ordinary least square (OLS) regression. The linear combination of organizational measure is significantly related to employees’ computer self-efficacy measure, F(6, 345)=5.078, p < .00. The standardized regression coefficients are shown in Table 4.



Table 2. Rotated Component Matrix for Organization Culture

Teamwork

Climate & Morale

Information Flow

Involvement

Supervision

Meetings

1 2 3 4 5 6

.001 .132 .157 .126 .276 .127

.138 .007 .345 .323 .284 .008

.671 .871 .526 .506 .623 .880

.202 .009 .231 .240 .120 .105

.208 .006 .360 .220 -.005 .001

.189 .007 .002 .003 .006 .007

7 8 9 10 11

.170 .252 .193 .193 .191

.515 .722 .823 .716 .823

.231 .196 .145 .332 .145

.243 .009 .179 .008 .179

.140 .164 .260 .003 .260

.411 .203 .119 .296 .119

12 13 14 15

.300 .210 .192 .281

.151 .355 .222 .118

.212 .163 .122 -.006

.001 .196 .186 .007

.674 .650 .756 .798

.330 .124 .261 .224

16 17 18 19

.005 .145 .410 .126

.398 .164 .268 .006

.188 .241 -.008 .005

.154 .162 .123 .120

.178 .276 .161 .189

.678 .627 .522 .727

20 21 22 23 24 25 26

.539 .778 .540 .561 .788 .741 .795

.263 .005 .155 .138 .197 .255 .009

-.009 .168 .131 .003 .183 .177 .199

.149 .001 .189 .162 .159 .131 .161

.159 .149 .235 .392 .187 .136 .106

.443 .196 .375 .009 .005 -.003 .007

27 28 29 30 31

.237 .154 .004 .129 .208

.188 .004 .129 .125 .204

.209 .009 .112 .179 .223

.512 .757 .822 .830 .667

.308 .244 .004 .004 .210

.116 .002 .194 .174 .008

Cronbach’s

.8844

.8487

.8619

.9062

.8781

.7741

Item

Extraction Method: Principal Component Analysis Rotation Method: Varimax with Kaiser Normalization

According to the regression coefficients, teamwork and information flow have positive and significant association with employees’ computer self-efficacy. Thus, Hypotheses 1 and 4 are supported. This implies that teamwork and information flow affects an employee’s computer self-efficacy. Under a supportive teamwork environment with good information flow, employees get encouragement and feedback from other employees, watch their colleagues use computers successfully and easily collect other related information to dynamically adjust their judgment toward their ability to use computer technology. Copyright © 2004, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea Group Inc. is prohibited.


Table 3. Pearson Correlations Among Variables

Climate & Morale

Team- Climate & work Morale .57 (.00)

Infoflow

Involvement

Supervision

Meetings

Infoflow

.43 (.00)

.59 (.00)

Involvement

.41 (.00)

.62 (.00)

.62 (.00)

Supervision

.45 (.00)

.57 (.00)

.61 (.00)

.59 (.00)

Meetings

.50 (.00)

.49 (.00)

.44 (.00)

.46 (.00)

.47 (.00)

Efficacy

.23 (.00)

.13 (.01)

.15 (.01)

.01 (.82)

.06 (.30)

.13 (.01)

Cronbach Alpha

.86

.91

.88

.78

.89

.85

Efficacy

.93

Note: p values for the significance of correlation are reported in parentheses

Table 4. OLS Regression Coefficients Predictors

Coefficients(E)

Teamwork

.214*** (.00)

Climate and Morale

.05 (.50)

Information Flow

.18** (.02)

Involvement

-.18** (.01)

Supervision

-.09 (.20) .05 (.43)

Meetings

* p