In this chapter, research framework and research design are described. Procedures and hypotheses are introduced in the research framework section. The method of sampling and data collection and measurements of each variable are illustrated in research design section.
Research Framework
According to the previous literature review, the possible antecedents were inspected by adopting the technology-organization-environment framework. This framework is applied for investigating innovation adoption at the organizational level, thus, the information technology capability, strategic leadership, and competitive tension are selected as the antecedents to examine the practices of e-HR adoption in the research. These three components are the independent variables as declared in the prior chapter, while organizational outcomes are the dependent variables to test the possible consequences of e-HR practice usage.
TabIn addition, as mentioned in the previous chapter, organization size and industry sector are factors that may impact firm performance. Thus, they are designed as control variables in the research framework.
Figure 3.1. shows the research structure of this study.
28
Figure 3.1. Research framework
The quantitative method was adopted to test the direct effects of competitive tension, strategic leadership, and IT capability on practices of e-HR and also to examine the effects of e-HR on organizational outcomes. A questionnaire was designed to collect data on participating organizations’ strategic leadership, IT capability, and practices of e-HR. To minimize the effect of common method variance, data on competitive tension and organizational outcomes were compiled from secondary information sources.
29
Research Hypotheses
Based on the previous literature review, the hypotheses of this research are as followings:
Hypothesis 1: Competitive tension of the firm in the industry has a positive influence on the practices of e-HR.
Hypothesis 2: Strategic leadership of the firm has a positive influence on the practices of e-HR.
Hypothesis 3: IT capability of the firm has a positive influence on the practices of e-HR.
Hypothesis 4: Practices of e-HR has a positive influence on the organizational outcomes.
Hypothesis 4a: Practices of e-HR has a positive influence on the HR efficiency.
Hypothesis 4b: Practices of e-HR has a positive influence on the ROA of a firm.
Hypothesis 4c: Practices of e-HR has a positive influence on the EPS of a firm.
Research Procedure
The first step of the research procedure was developing the research topic. After reviewing HR literatures, the researcher narrowed down the research topic to investigate the e-HR adoption and organizational outcomes in Taiwan. Second, through reviewing the relevant literatures, the researcher determined the questions that should be answered in this study. Third, the researcher continued reviewing relevant literature to get better understanding of the e-HR field. Other review focuses were strategic leadership, IT capability, competitive tension, practices of e-HR, and organizational outcomes. Forth, the research framework was drawn and instruments were selected to examine study variables. Next, in the stage of data collection, publicly listed companies in Taiwan were selected as the sample to
30
test the research framework. Once data were collected, SPSS and SmartPLS were adopted to analyze the data. The final step included the formation of conclusions and discussion; the outcomes of hypotheses testing were interpreted and research questions were answered. The procedure of this research is shown below in Figure 3.2.
Figure 3.2. Research procedure
Research Design
This research adopted a quantitative survey method to examine the hypotheses. The publicly listed companies in Taiwan were the sample of this study, and a census approach was applied for collecting data. The HR professionals of listed companies were the respondents to answer the questionnaire. The measurements of variables went through exploratory factor analysis (EFA) and confirmatory factory analysis (CFA) to ensure the validity and reliability.
31
Sample and Data Collection
The sample for this study targeted at the publicly listed companies in Taiwan and the sampling frame was the companies listed in the Taiwan Stock Exchange to ensure accessibility of organizational performance data. According to the public report from the Taiwan Stock Exchange, there were 843 listed companies in 2012 (Taiwan Stock Exchange, 2012). Data availability and accuracy were main considerations for the researcher to choose companies listed in the Taiwan Stock Exchange.
A census-approach was implemented to collect data from all companies in the sampling frame, and mail survey was the main instrument to gather information. The questionnaires were mailed to the HR department of all listed companies in April 2013, with a cover letter that explained the purpose of this research and a coupon in the value of NT$100 for appreciation and it took one month to gather the responses. To boost the responses, the researcher also attended campus recruitment activities to solicit participation from HR personnel at each company booth. Personal networks were also tapped to increase the response rate.
HR managers and professionals of listed companies were targeted to fill out the survey.
If there were multiple copies of the survey completed for one company, the researcher used the following criteria to select the most representative response: 1) human resource working experience, 2) seniority in the company, 3) functional content, and 4) computer literacy.
Sample Profile
A total of 258 responses were collected, and 204 responses were kept to do data analysis after screening out invalid data. The valid response rate of census approach was 16%, while other 79 valid responses were gathered from campus recruitment activities and personal networks.
To briefly describe the distribution of total 204 participants, 68.6% of them were
32
females, while the male accounted for 30.4%. In the age segmentation, most of the participants ranged from 26 to 35 years old with 42.6% and 40.2% of them ranged from 36 to 45 years old. As for the seniority in company, 40.2% of the participants had 1 to 5 years experiences, and followed by 19.6% of the participants with 6 to 10 years experiences in the same company, the others accounted for the rest 40.2%. In the section of human resource working experience, 34.3% of the participants had 1 to 5 years in HR related work, 26.0% of the participants had 6 to 10 years in HR related work, and 21.2% of the participants had 6 to 10 years in HR related work; the others accounted for the rest 18.5%. In general, participants are confident of their computer skills, the majority of the participants rated themselves good as to the computer skills; the other 31.4% rated themselves as normal. As for the functional expertise, 31.9% of the HR professionals perform at least three functions, 25% of the HR professionals perform at other functions like general management office or personnel office, and 14.2% of the HR professionals perform at recruiting function; the others functions accounted for 28.9%. Sample information regarding the research respondents are reported in the following tables. Table 3.1 shows the personal information toward the respondents.
Table 3.1.
Descriptive Information on the Respondents
Sample characteristics Frequency Percentage
Gender 1. Male 62 30.4%
2. Female 140 68.6%
Missing value 2 1%
Age 1. 25 and below 3 1.5%
2. 26-35 87 42.6%
3. 36-45 82 40.2%
4. 46-55 27 13.2%
5. 56-65 5 2.5%
Missing value 0 0%
(continued)
33
Table 3.1 (continued)
Sample characteristics Frequency Percentage
Company seniority 1. Below1 year 17 8.3%
2. 1 year to 5 years 82 40.2%
3. 6 years to 10 years 40 19.6%
4. 11 years to 15 years 33 16.2%
5. 16 years to 20 years 12 5.9%
6. 21 years and above 20 9.8%
Missing value 0 0%
Human Resource seniority
1. Below 1 year 6 2.9%
2. 1 year to 5 years 70 34.3%
3. 6 years to 10 years 53 26.0%
4. 11 years to 15 years 43 21.1%
5. 16 years to 20 years 22 10.8%
6. 21 years and above 10 4.9%
Missing value 0 0%
Computer skill 1. Very poor 1 0.5%
2. Poor 3 1.5%
3. Average 64 31.4%
4. Good 119 58.3%
5. Excellent 16 7.8%
Missing value 1 0.5%
Functional department
Recruiting 29 14.2%
Training 15 7.4%
Compensation and benefit 16 7.8%
Other functions 57 28.0%
Two functions 17 8.3%
Multiple functions 65 31.9%
Missing value 5 2.5%
In addition, respondents were from different industries. The majority of the respondents were from manufacturing industries. More detailed information was organized in Table 3.2.
34
Table 3.2.
Descriptive Information Regarding the Participating Companies
Sample characteristics Frequency Percentage
Industry
Manufacturing 145 71.1%
Service 23 11.3%
Others 32 15.7%
Manufacturing and Service 1 0.5%
Missing value 3 1.5%
Participants were also asked to answer the year that company implemented integrated software. Most of the participants did not answer this item due to a lack of knowledge of the specific time of implementation, thus, the missing value accounted for the most proportion.
The distribution of implementing year was organized in Table 3.3, and it showed most of the
35
sample company implemented integrated software after the year 2000.
Table 3.3.
Year of Implementing Integrated Software
Year Frequency Percent
missing value 76 37.3
1990 1 0.5
1995 2 1.0
2000 10 4.9
2001 8 3.9
2002 1 0.5
2003 6 2.9
2004 5 2.5
2005 4 2.0
2006 3 1.5
2007 7 3.4
2008 10 4.9
2009 2 1.0
2010 10 4.9
2011 9 4.4
2012 1 0.5
2013 3 1.5
long time ago 1 0.5
Company did not implement integrated software 45 22.1
Total 204 100.0
Questionnaire Design
The portion of the questionnaire related to the current study was divided into four parts to collect data on IT capability, strategic leadership, practices of e-HR, and demographic information. The HR practitioners were the sample to fill out the questionnaire. Taking common method variance issue into consideration, data of competitive tension and organization outcomes were acquired from secondary information.
The first part of questionnaire is IT capability which included two dimensions and 13
36
items. Moreover, three reverse coded items were applied to screen out invalid responses.
The items adopted the 7-point Likert scale from strongly disagree (1) to strongly agree (7), and participants were guided to answer questions based on their interaction with IT department. The second part is strategic leadership which included 12 items; the visionary leadership and strategic execution of high-level leader in company were examined. The items adopted the 7-point Likert scale from strongly disagree (1) to strongly agree (7) to indicate the level of strategic leadership of the company leader. The third part is the scale for e-HR practices. There are 51 items of HR activities and participants were guided to answer each item based on their actual work practice. The scale ranged from 1 to 5 to show different degree of automation. The last part of the questionnaire contains demographic questions, participants were asked to answer age, company seniority, HR seniority, computer skill, functional department, and other company information such as the number of HR personnel, employee number, capital, and industry sector.
Since the study was conducted in Taiwan, the questionnaire adopted the Chinese translation version from a previous study to make it easier to understand by the participants.
The Chinese translation version had gone through expert review to ensure the validity of its content. The questionnaire can be seen in Appendix A.
Measurement
This section reports the measures used in this study, source of these measures, as well as the analysis and results of validity and reliability testing of latent measures. In order to validate major variables, exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were applied to examine the validity, and Cronbach’s alpha for reliability respectively.
Furthermore, the secondary information was acquired from the Taiwan Economic Journal (TEJ) Database, which was founded in 1990, and keeps updated public financial report that
37
companies disclose to meet legal requirements. The TEJ database is widely used by companies or universities to collect various financial reports.
Competitive Tension
Competitive tension aimed to measure the level of competition within an industry.
Carroll and Hannan (1992) posited that new company is hard to survive in industry with lots of existing companies, and the number of company has negative influence on the establishment of new company. Thus, the competitive tension of company was measured by the total number of companies in the same industry category in this study. The higher the number of companies in one industry category, the higher the competitive tension a company faces.
The classification of industry sector organized by the Taiwan Stock Exchange (TSE) was adopted, and the information was gathered from the TEJ database. The detail classification was reported in Appendix B.
Strategic Leadership
The study adopted the scale developed by Chen and Wu (2008) as the measurement of strategic leadership. Chen and Wu developed and validated the measurement through different tests included reliability, discriminant validity, convergent validity, and nomological validity. Then, the pilot test was conducted for gathering data to do exploratory factor analysis (EFA), confirmatory factor analysis (CFA), and reliability analysis.
Two dimensions of strategic leadership were applied as the instrument for this study.
These two were visionary leadership with a published Cronbach’s alpha of 0.92 and strategic execution of 0.90. These provided evidence of a well-developed instrument for measuring strategic leadership. Each dimension consisted of 6 items. One sample item for visionary leadership was “Manager draws the blueprint for our team.”, and for strategic execution was
“Manager builds relationships inside and outside the organization.”
38
IT Capability
IT capability in this study aimed to measure the ability of IT department to cooperate with HR practitioners. This measure was based on Goodhue and Thompson’s research in 1995, which proposed and constructed the task-technology fit theory based on users’
perspective.
The validity was tested later by Goodhue in 1998 based on users’ evaluation of information system. After testing the reliability and discriminant validity of the items, the measurement was reduced from 48 to 34 items. This measurement included eight components:
quality, locatability, authorization, compatibility, ease of use/training, production timeliness, systems reliability, and relationship with users to measure how well these characteristics of technology fit the tasks of the users.
The components of systems reliability and relationship with users were adopted in this research because these factors were more appropriate for accessing IT capability or the ability of IT to support other functional areas in the company. The published Cronbach’s alphas of each dimension were 0.71, and 0.88.
A sample item for systems reliability is: “I can count on the system to be "up" and available when I need it”. And the sample item for relationship with users is: “The IS people we deal with understand the day-to-day objectives of my work group and its mission within our company”. A complete list of items can be found in Appendix C.
Practices of e-HR
Items related to practices of e-HR were intended to gauge how widely the HR staff uses e-HR functions. The measurement of practices of e-HR adopted the comprehensive list of e-HR practices that was developed by Haines and Lafleur in 2008 after reviewing professional literature and vendor package. There were 78 applications in human resource activities and covered nine dimensions: HR audits and survey, employee benefits,
39
compensation and rewards, health and safety, performance management, HR planning and career development, staffing, training and development, and employee relations.
In addition, Yeh and Wei (2011) translated Haines and Lafleur’s measurement for companies in Taiwan and validated the Chinese version measurement through focus group interview, expert review, and pilot test to make sure the validity and reliability of this measurement. Two dimensions were deleted after the validation process, the HR audits and survey, and the health and safety. Thus, seven dimensions were left and formed a new measurement. Hence, this study adopted the Yeh and Wei’s (2011) version of the measurement to investigate the practices of e-HR usage in Taiwan. The revised measurement and entire items were listed in Appendix C.
This scale ranged from 1 to 5 to represent the usage of e-HR practices. The higher the score is, the higher the level of automation of an HR activity. The scale of 1 to 5 follow a progression from none to high level of automation, with 1 indicating the company does not have the activity, 2 the activity is carried out manually, 3 the activity involves the use of office software, 4 the activity involves the use of packaged software, and 5 the activity involves the use of integrated software.
A sample item is: “Allowing employees to access pay data information.” Participants were guided to answer from 1 to 5 of each practice based on their actual work experience.
When aggregating the score for e-HR practices from these activities, the values were recoded with no such practice as missing values, manual operation as 1, the use of office software as 2, the use of packaged software as 3, and the use of integrated software as 4.
Organization Outcomes
HR efficiency
The employee to HR personnel ratio (Watson Wyatt, 2002; Lenngnick-Hall & Moritz, 2003; Parry, 2011) was adopted as the measurement of HR efficiency. The item for HR
40
efficiency was asked by open-ended questions, the questions include “How many HR personnel are in your company?” and “How many employees are in your company?”
Since the number of employee was hugely different among companies, the logarithm of the numbers of HR personnel and employees were used in the calculation to better examine the impact of e-HR practices on the employee to HR personnel ratio. HR efficiency was calculated by the log value of employee number divided by the log value of HR personnel number. That is, the higher the value, the higher the HR efficiency of a company. The range of HR personnel number and employee number of participating companies is organized in the following Table 3.4.
Table 3.4.
Range of HR Personnel Number and Employee Number of Companies
N Min. Max. Mean S. D.
Employee number 203 5 95000 3590.09 9580.95
HR personnel 193 1 400 15.41 36.020
Financial performance
Financial performance was applied as the dependent variable many times in various research fields. To measure the influence of e-HR use on organization financial performance, return on asset (ROA) and earnings per share (EPS) were used as two measures to examine the financial result. ROA and EPS were two of the most important financial indices that publicly listed companies need to disclose.
The information of ROA and EPS were collected from the TEJ database, hence, survey participants did not have to answer these items.
41
Control variables
The control variables were included to control variance that might cause estimation bias.
Two control variables of company background were applied in this research.
The first control variable is the size of an organization. Firm size is viewed as a source of organizational cost (Shepherd. 1972). In addition, scholars posited that firm size is an indicator of the diversification of an organization based on the strategy perspective (Rumelt, 1982; Porter, 1987). Thus, organization size can be one of the general factors that influence the cost and strategy selection of a firm. The organization size was measured by the company capital; the information was provided by participants or acquired from Taiwan Stock Exchange. The log value of capital was used due to a substantially wide range of company capital in the study sample.
The second control variable is the industry sector. Schmalensee (1985) stated that the majority of variances among business unit performances could be explained by different industry sector. Thus, industry sector can be another factor to influence the performance of firms among different industries. In this study, two main industries were selected, manufacturing and service, to separate participants’ companies generally. However, most of the sample belonged to the manufacturing industry. Hence, manufacturing industry and non-manufacturing industry were coded in a dummy variable to control the industrial variance.
Validity and Reliability Testing for Measurement
The latent measures in this study, namely strategic leadership, IT capability, and e-HR practices, went through several statistical analysis procedure to ensure their validity and reliability. These procedures included exploratory factor analysis, confirmatory factor analysis, and internal consistency reliability testing. SPSS was used to perform exploratory factor analyses and internal consistency reliability testing using Cronbach’s alpha. SmartPLS
42
was used to perform confirmatory factor analyses.
After data collection, exploratory factor analysis (EFA) was performed on the measurements of strategic leadership, IT capability, and e-HR practices for two purposes: 1) to detect the problem of common method variance through Harman’s one factor test, and 2) to ensure that factor structure of the data corresponded to the original measurement.
Harman’s one factor test in SPSS was run with principal components extraction method and varimax rotation. The value of KMO measure of sampling adequacy was 0.86 which
Harman’s one factor test in SPSS was run with principal components extraction method and varimax rotation. The value of KMO measure of sampling adequacy was 0.86 which