• 沒有找到結果。

The subsequent chapter outlines the research methodology. It includes the research framework, states the hypothesis tested by the researcher, depicts the research procedures that were followed, describes the research design and ends with an explanation of the statistical analysis methods through which the empirical data obtained was evaluated.

Research Framework

The TIP Model (which stands for Trust, Intellectual Capital and Performance) shown in figure 3.1 was developed by Shih and Funes based on the Model of Organizational Trust, Job Satisfaction and Effectiveness of Shockley-Zalabak, Ellis and Cesaria (2003), Scandia’s Navigator (Edvinsson, 1997) and the Survival and Competitiveness of the Enterprise Intellectual Capital Model (Lin, 2012). The TIP model served as the research framework for this study. The framework clearly depicts the hypothesis tested and the variables studied by the researcher.

Employee

· Employees with key knowledge

· Employees learning and update speed H1

· Excellent and distinctive culture and value system Environment

· Adapt to environmental capacity

· Good working environment

· Good relationships of costumers, suppliers and partners

23

Research Hypotheses

Based on the research questions, and the literature review, the following null-hypotheses were formulated as follow:

H1: Organizational trust has no effect on intellectual capital.

H2: Intellectual capital has no effect on business performance.

H3: Business performance has no effect on survival and competitiveness.

H4: Innovation capital has no effect on structure capital.

H5: Process capital has no effect on structure capital.

H6: Customer capital has no effect on intellectual capital.

H7: Structure capital has no effect on intellectual capital.

24

Research Procedure

The research was conducted following the subsequent research process:

Figure 3.2. Research Process

Research Design

Because this study is a survey correlational research, a quantitative approach was implemented. This methodology was selected because it was the intent of the researcher to collect numerical data which was then analyzed using mathematically based methods with the purpose of testing whether or not there was a statistical relationship between the variables of interest (Sukamolson, 1996). In addition, a quantitative method approach facilitates the description, illustration and exploration of a given phenomenon (Yin, 2003), therefore it is deemed appropriate for the purposes of this research.

Report research Findings and Conclusions Data Analysis

Gather Data Conduct Pilot Study

Translation, Backward-translation and Expert Review of instrument Instrument Development

Develop Research Method of the Study Develop Theoretical Framework of the study

Review of Literature

Identify Research Questions and Hypothesis

25

Data Collection

The target population of this research are employees of Ficohsa Financial Institution.

Because of this institution’s relevant position within the Honduran financial system, it was considered appropriate by the researcher to use for this investigation. To gather the necessary data, an online questionnaire was utilized. The instrument was first tested on a group of 30 Ficohsa employees who participated in a pilot study. Once the instrument was revised and validated, it was then implemented to gather data for the main study.

To date, Ficohsa employs 1,200 people; therefore, the minimum sample size in order to be able to make generalizations about the entire population from the results obtained is of 291 individuals. This number was acquired by utilizing the sample size formula for an infinite population and then using the sample size derived from that calculation to calculate the sample size formula for finite population. This procedure is followed to calculate the minimum sample size of a population less than 50,000 as is the case with this organization.

(Godden, 2004):

( ) ( )

Where SS= Sample size; Z= Z-Value; p= percentage of population picking a choice, expressed as a decimal and C= confidence interval, expressed as a decimal. For the purpose of this study the Z-value, which represents the probability that a sample will fall within a certain distribution, was 1.96 because a 95% confidence level was used, P= .5 because this is the number used when determining sample size needed and C= .05 because for this research a Standard Error of 5% is considered acceptable.

( )

( ( )) Where Pop= population.

26 investigation were randomly chosen. Employees are assigned an employee identification number when they first join the organization. In order to select participants for the research a web based random number generator, Random.com was used to obtain 330 numbers (30 numbers for the pilot study and 300 for the main study). Numbers obtained were then sent to the HR manager who matched them with the corresponding employee identification numbers.

The HR manager then contacted participants using the company’s intranet requesting their participation in the study. After they agreed, the HR manager emailed respondents a link to access Kwiksurvey, an online survey builder, were the questionnaire was hosted. This method was chosen because of its convenience and low cost. All ethical guidelines as well as the confidentiality of all participants were strictly upheld during the course of this research

Instrument

The instrument consists of 8 variables and has a total amount of 70 questions that are divided into four distinct parts, these are: part I) organizational trust; part II) intellectual capital; part III) business performance and survival and competitiveness; and part IV) demographics. Parts I through III were evaluated by using a 5 point Likert-type Scale asking participants to rate each statement based on their opinion about their company (1=strongly disagree, 2=disagree, 3=average, 4=agree,5=strongly agree). For part IV, respondents were asked to choose one of different available options.

Part I, which relates to organizational trust, was adapted from the 29-item Organizational Trust Index (OTI) developed by Pamela Shockley-Zalaback, Donald Morley, Ruggero Cesaria and Kathleen Ellis. The index was developed by making normative comparisons from data obtained from a total of 53 organizations from the United States (25 states), Italy (11 cities), Sydney, Singapore, Hong Kong, Tokyo, Bombay and Taiwan. This data expressed the opinions of a total of 4,000 supervisory and nonsupervisory employees from multiple industries including banking, telecommunications, manufacturing, IT, sales and customer service. The validity and reliability of the 29-item OTI was sufficiently established in the author’s original work.

27

Part II was adapted from questionnaires relating to intellectual capital and management systems by Cabrita & Bontis (2008), Debowski (2006), Bukowitz & Petrash (1997), Black & Lynch (2005) and Bontis (2004). A total of three questions per variable were selected from the different questionnaires adding up to a total of 18 items. The validity and reliability of each one of the above mentioned instruments were also sufficiently established in the author’s original work.

Part III explores two dimensions: survival and competitiveness and business performance. The 9 items for survival and competitiveness were adapted from the work of Lin (2012) on knowledge management. Primarily, the questions used in this section are based on the literature discussing the factors that mainly affect the survival and competitiveness of an enterprise. Finally, business performance is assessed through market leadership and financial performance. The 10 items for this section (5 for market leadership and 5 for financial performance) were adapted from the empirical research of Lee and Choi (2003).

The validity and reliability of the items, as established in the authors’ original work, are suitable for this research.

Construct Coding and Scales

Tables 3.1 thru 3.4 show the items used to measure the variables of this study. Each item was given a code that was later used in the statistical analysis of the data. Dummy variables were created to code part IV of the measurement instrument pertaining to demographics.

Table 3.1

Items measuring Organizational Trust

Construct Code Questionnaire Item

Competence Comp1 Overall operational efficiency.

(Comp; 4 items) Comp2 Overall quality of financial products and/or services offered to clients.

Comp3 Capacity to achieve objectives.

Comp4 Capability of employees.

Openness/Honesty (Op/h; 9 items)

Op/h1 Honesty with immediate supervisor when things are going wrong.

Op/h2 Freedom to disagree with immediate supervisor.

(continued)

28

Construct Code Questionnaire Item Op/h3 Say in decisions that affect job.

Op/h4 Ability of immediate supervisor to keep confidences.

Op/h5 Reception of adequate information regarding job performance.

Op/h6 Reception of adequate information regarding performance evaluation.

Op/h7 Reception of adequate information about how job-related problems are handled.

Op/h8 Reception of adequate information regarding how organizational decisions that affect job performance are made.

Op/h9 Reception of adequate information regarding long-term organizational strategies.

Concern for employees (Conc; 7 items)

Conc1 Willingness of immediate supervisor to listen to employees.

Conc2 Sincerity of top management in their effort to communicate with employees.

Conc3 Willingness of top management to listen to employee’s concerns.

Conc4 Concern from immediate supervisor for employee’s personal well-being.

Conc5 Concern from top management for employee’s well-being.

Conc6 Sincerity of immediate supervisor in his efforts to communicate with team members.

Conc7 Supervisors speak positively about subordinates in front of others.

Reliability (Reli; 4 items)

Reli1 Follow through by immediate supervisor on what he says.

Reli2 Consistent behavior from immediate supervisor.

(continued)

Table 3.1 (continued)

(continued)

29

Iden3 Connection to immediate supervisor.

Iden4 Similarity in values to peers.

Iden5 Similarity in values to immediate supervisor.

Table 3.2

Items measuring Intellectual Capital

Construct Code Questionnaire Items

Human Capital

OrgC1 Variety of training (formal and informal).

OrgC2 Formal and effective grievance procedure or complaint resolution system.

OrgC3 Flexible organizational structure.

Customer Capital (CuC; 3 items)

CuC1 Customer satisfaction.

CuC2 Circulation and understanding of customer’s feedback and comments.

CuC3 High-value added customer service.

Structure Capital (StruC; 3 items)

StruC1 Knowledge management system improves job performance.

StruC2 Regular update and ability of knowledge management system to provide real-time information.

Table 3.1 (continued)

(continued)

30

Construct Code Questionnaire Item

StruC3 Knowledge management system meets user’s needs.

Innovation Capital (InnC;3 items)

InnC1 Open-minded culture that supports innovation.

InnC2 Knowledge sharing.

InnC3 Management style supports and encourages innovation.

Process Capital (ProC;3 items)

ProC1 Taking full advantage of information and communication technology to facilitate the transfer of information.

ProC2 Software allows easily processing, store, retrieving and communicating information.

ProC3 Hardware adequately supports software.

Table 3.3

Finp1 After tax return on sales.

Finp2 Profit growth.

Finp3 After tax return on assets.

Finp4 Average profit.

Finp5 Overall business performance.

Table 3.2 (continued)

31 Table 3.4

Items measuring Survival and Competitiveness.

Construct Code Questionnaire Items

Employee

(SuCemp; 2 items)

SuCemp1 Recruitment and retention of employees with key knowledge.

SuCemp2 Employee learning and update speed.

Organization (SuCo;4 items)

SuCo1 Innovation capability.

SuCo2 Flexible organizational structure.

SuCo3 Dynamism.

SuCo4 Distinctive culture and value system Environment

(SuCen;3 items)

SuCen1 Ability to easily adapt to the environment’s capacity.

SuCen2 Work environment

SuCen3 Relationship between customers, suppliers and partners.

Translation and Face Validity

The questionnaire was translated from English to Spanish by the researcher. The translation was then revised by two bilingual financial experts. Finally, the questionnaire was submitted to a backwards translation to ensure the original meaning of the items was not lost in the translation process. For the convenience of respondents, the instrument stated each question in both languages.

Data Analysis Validity and Reliability

To test the validity of the instrument, face validity was employed. Face validity refers to the “simple form of validity, in which researchers determine if the test seems to measure what is intended to measure by looking at whether a test appears to measure the target variable” (Anastasi, 1988). The English version of the instrument was validated by one HRD academic professional, the Spanish version on the other hand, was validated by one non-academic HRM professional and two non-non-academic financial professionals.

To test the reliability of the measures, Cronbach’s Alpha and individual item reliability (i.e. factor loading) were used. The coefficient alpha score assesses the internal

32

consistency of the questionnaire. Nunnally (1978) states a reliability value of .7 is the minimum reliability for research. According to Hair et al. (2006) individual item reliability is confirmed when item loading exceed the factor loading cutoff point according to the sample size. Hair et al. guidelines for factor loading cutoff points based on sample size are presented in the table below.

Table 3.5

Hair et al. (2006) guidelines for factor loading cutoff points based on sample size Sample Size needed for significance Factor Loading

350 .30

250 .35

200 .40

150 .45

120 .50

100 .55

85 .60

70 .65

60 .70

50 .75

Construct validity was established through Explanatory Factor Analysis (EFA). EFA was used to test if the factors of each scale were consistent with those of previous studies.

Methods of Data Analysis

The empirical data collected from the sample was evaluated using the following methods:

Descriptive Statistics

Descriptive statistics were used in order to obtain a general description of the basic features of the data. This holistic overview provides a simple summary about the sample in order to have a better understanding of where the data came from.

33 Correlation Analysis

A correlation analysis was performed to obtain a correlation matrix. The matrix was then examined in order to detect if multicollinearity was present in the model.

Multicollinearity occurs when two variables are closely correlated to one another. This indicates they are likely to be measuring the same construct. SPSS version 20 was used to perform this analysis.

Partial Least Square (PLS)

The empirical data collected from the sample population was evaluated through Partial Least Square (PLS). Partial Least Square (PLS) path-modeling algorithm is a type of structural equation modeling technique. It was developed by Herman Wold in 1975. PLS estimates path models using latent variables. It allows the simultaneous modeling of relationships among multiple constructs; in other words it permits the analysis of a system of constructs. The objective in PLS is to “maximize the explanation variance, thus R2 and the significance of relationships among constructs are measures indicative of how well a model is performing” (Bontis, 1998, p. 69). A PLS model consists of a structural part which reflects the relationship between the variables and a measurement component which shows how the variables are related. SmartPLS 2.0 software will be employed to perform a confirmatory factor analysis and a path analysis.

Before testing the structural model, reflective measurements were obtained through Smart PLS 2.0 in order to ensure the internal consistency reliability, indicator reliability and KMO (Hair et al., 2006). The measurement used to test the internal consistency reliability is the composite reliability. In exploratory research, this value should be higher than .7 in order to be deemed acceptable. Indicator reliability is measured by their loadings. The loadings prove the validity of each of the indicators and measure how much of the variance is explained by that specific variable. KMO is the acronym for the Kaiser-Meyer-Olkin Measure of Sampling Adequacy. This is a statistic that indicates the proportion of variance in the variables that might be caused by underlying factors. High values, these are those closer to 1.0, generally indicate the data is likely to factor well. A value of .5 is considered acceptable to proceed with factor analysis (Garson, 2013).

After the model was successfully validated by the above mentioned tests, further analysis of the structural model were made by evaluating the coefficient of determination (R2), path coefficients, t-value (t) and bootstrapping. R square is used to explain the endogenous latent variables to the total variance. Chin (1998) proposed R2 values of .67, .33 and .19 are considered substantial, moderate and weak respectively. Path coefficients are used to judge

34

the relationship between variables, determine the direction of the relationships and its significance. In SmartPLS this value can be determined by bootstrapping the data and examining the resulting t-value. Heir et. al. (2006) defines the critical values in a two-tailed test to be weak when they are less than 1.65, moderate when they are between 1.65 and 1.96 and strong when they are equal or greater than 2.58.

Bootstrapping is “a nonparametric approach to statistical inference that does not make any distributional assumptions of the parameters like traditional methods. It draws conclusions of a population strictly from the sample at hand” (Sharma & Kim, 2012, p.2).

This is estimated by resampling from a sample using replacement. Each re-sample has the same number of elements as the original sample while “the replacement method ensures each re-sample is slightly and randomly different than the original sample” (Sharma & Kim, 2012, p.2) allowing the researcher to create a simulated large sample and estimate coefficients that can be tested for their significance.

35

CHAPTER IV DESCRIPTIVE STATISTICS FINDIGNS

相關文件