Chapter Overview
The chapter explains how the research framework for patterns of winning by Intellectual Capital will be tested, objective of this chapter are seeks to illustrate the research methodology consist of the conceptual model in research framework, followed by the research hypothesis, a glance look of research procedure, data collection, measurement instrument, and validity and reliability of the study, to provide audience a clear navigation about how the study will be conduct, and last but not least, the methods of data analysis will be then provided.
Research Framework
As discussed previously, the research model for patterns of winning by Intellectual Capital will be based on the literature review, to build a comprehensive research model, which consists of knowledge management strategies, knowledge enabler, intellectual capital, and business performance. This study provides an integrated intellectual capital model named as the KlPs Model, which is an abbreviation, the K stands for Knowledge Management strategies and Knowledge enabler, the I stands for Intellectual capital, and Ps stands for Business Performance, the model was specially developed by Mingchu Aibo Tsai and Cheng-Ping Shih for this study, to measure the effect of each constructs with one another by the statistical parameter, seek to understand the genetic cause and investigation on the winning patterns by the Intellectual Capital, the key Knowledge Enabler, most effective Knowledge Management Strategies for Taiwan financial institution to cope with FinTech challenges by and large.
The model was based on several previous researchers study, such as Knowledge Management Strategies of Choi and Lee (2002), Knowledge Enabler of Paul (2011), Intellectual Capital of Bontis (1998), and Business Performance of Chou (2012).
Figure 3.1. Research framework
Research Hypotheses
Intellectual capital plays a central role in business survival and performance (Chen et al., 2004; Enz, Canina and Walsh, 2006). Academics have traditionally been very interested in how intangible assets reflect the performance of firms (Chenhall and Smith, 2007). Based on the framework revealed in figure 3.1 and on the basis of the previous research questions, there are three hypotheses assumed in a numerical order and stated as follows.
Hypothesis 1: Knowledge Management Strategies has no effect on Knowledge Enabler.
Hypothesis 2: Knowledge Enabler has no effect on Intellectual Capital.
Hypothesis 3: Intellectual Capital has no effect on Business Performance.
Research Procedure
This research is conducted followed by the procedure shown below.
Figure 3.2. Research procedure
Data Collection
This study aims to understand how traditional financial institution can best cope with FinTech challenges, by examine the knowledge management strategies, knowledge enabler, intellectual capital on business performance, the following section will discussed about the participants of the study, population of the study, measure instrument, validity and reliability and method of data analysis to be discussed here.
Participants
The sample participants are the employees worked in finance related industry present as well as past. Due to the large number of participant can fall into the category, this study aims in study the financial institution, majorly the banking sector, with most of the market presence, in another word, the sample participant’s branch can be seem throughout Taiwan, to provide a vivid and comprehensive picture of the current winning patters by intellectual capital in financial institution in Taiwan, in understand the significant factor as well as drive for the modern knowledge-intensified organization.
The sample are randomly assigned in branches throughout Taiwan, aims in provide the most unbiased result, despite of geographic location in Taiwan, to use one bank which has branches all around Taiwan, seek to depict the picture of current banking sector.
Population
The population for the pilot study consists of 47 valid samples out of the total sample collected after deleting invalid samples, and for the main study, sample is concentrated in one of the largest Taiwanese bank, which has some 160 branches all around Taiwan, provide financial services from corporate banking, personal banking, as well as investment services, with the long history of serving clients of all kinds, which the researcher in this study think is a perfect sample to Study and represent the current winning patterns of intellectual capital initiative in Taiwan banking sector. The study try to generate unbiased result by distributing the participants located in branches around Taiwan. However, the respondents are still with gender bias, which female employees are taking large pool of the sample and it may due to two reason. First, the female employees are still the largest group of employees in terms of banking sector.
Almost every bank branches in Taiwan, are consists of large group of female employees, which it seems to be the banking industry prefer or predominately by the female employees. Second, the large pool of female employees may due to the study
sample focused on banking sector, and mostly from the branches level, if the study was conducted in a corporate level, with a more specific focused on certain category of job roles as well as job responsibilities, such as finance, investment, wealth management, human resource and strategic planning, the result may be more fairly distributed.
The current population for this study are consists mostly employees with working experience in the financial related industry for approximately 6 to 10 years followed by employees with 3 to 5 years of experience, with the position of assistant / associate role, followed by the management trainees / associates position, are majorly accounted for this study, which the finding from this study may be somewhat more helpful in understand for an associate / assistant as well as management trainee / associates position employee’s status quo, rather than other categories.
However, this study trying to overcome the possible gender as well as position based bias, the research adopted the partial least square using structural equation model tested by the statistical software “SmartPLS” version 3.2.4, using the bootstrapping method to duplicate data to cope with the nature of intellectual capital discussed in previous literature, which is difficult to duplicate, the study wish to use the statistical tool seek to minimize the issue, therefore, the study try to seek for a relatively more unbiased result by bootstrapping method, in understanding the winning patters of intellectual capital initiative.
Measure Instrument
The total of 60 questionnaire questions is designed to investigate the winning patters of intellectual capital with 57 questions from various construct and 3 other questions in examining the demographic. By making sure it is feasible to examine the responses, this study not only has professional experts reviewed, also confirmed with previous literatures in assuring the questionnaire which will be deployed are subject to validity and reliability in examine the winning patters of intellectual capital and strategic knowledge management in modern organizations.
1. Knowledge Management Strategies (8 items): Adopted from Hsieh (2007, pp.83) which there are two different view on orientation; one is the human-oriented strategy as the KMS H and the system-human-oriented are as KMS_S.
2. Knowledge Enabler (20 items): Adopted from Paul (2011) and only choose the four most important factor learning as KE_L, T-shaped as KE_TS, IT-support as KE_IT and Centralization as KE_C.
3. Intellectual Capital (20 items): Bontis (l998) developed the comprehensive instrument for measuring intellectual capital. The human capital refer to IC_H, structural capital refer to as IC_SC, relational capital refer to as IC_RC and innovation capital refer to as IC_IC.
4. Business Performance (9 items): Bontis (2010) developed the referable instrument. Please note, the business performance dimension, minor change has been made, by adding the measurement of customer satisfaction replacing the productivity perspective. Includes Market Value as BP_MV, and Financial Performance as BP_FP.
Last part of the questionnaire is designed to investigate the working experience, position in the company, and gender distribution. As each dimension is constructed, the questionnaire is also translated into Chinese for comprehension and confirmation.
All of the participants are advised to fill out the survey thoroughly. The questionnaire is designed and will be only filled by the respondents who work in the financial related industry, particularly in the leading banks. The responses of questionnaire are adopted a Five-point Likert-type scale ranging from one for strong disagreement to five for strong agreement. In the Appendix section, both the Chinese and English survey sample, and coding list respectively will be provided, for more information please refer to the appendix section.
Validity and Reliability
Validity of instrument is the degree to which scales measure what researchers claim they measure (Huot, 1990). The validity is determined by content validity, criterion validity and construct’s validity. In particular, the content validity is an extent to which the measuring instrument is used to provide the adequate questions of a study.
Reliability is the degree to which a scale is consistently reflected on the construct (Nahapiet and Ghoshal, 1998).
Reliability of instrument is the external and internal consistency of measurement.
Cronbach’s Alpha is used in this study, and is computed for each of the variables, measuring the dimensional evidence of reliability (Cortina, 1993) as well. Featured with the examination of internal consistency, the figure that is above 0.70 is considered acceptable (Guieford, 1965).
Again, the validity and reliability explanation as well as its implication in this study are classified in the following Table 3.1 and 3.2 respectively. Both validity and reliability of the instrument is determined by an internal consistency of the study, there are five expert complete assessments on each item of the questionnaire before conducting the survey, the purpose is to exploited and determine whether the items are appropriately designed and assigned. The explanatory factor analysis (EFA) is tested afterwards to see if the factors of each scale are consistent with previous studies.
Table 3.1. Validity of Instrument Validity of Instrument
Validity Type Theoretical Meaning Validity of Instrument Face
Validity
Validity is used as indicator to see whether it “Make sense” as a measurement in judging others, especially in the scientific community.
Measurement in representing all aspects of conceptual definition of a construct. explaining how well the indicators of single construct converge or how well the indicators of different construct diverge.
Using the KMO and Barlett’s test for Sphericity.
The following table 3.2 shows the reliability of instrument for its theoretical meaning, as well as the critical point in determine the reliability in an statistical sense, and introduced the measure used in this study for determine reliability.
Table 3.2. Reliability of Instrument Reliability of Instrument
Reliability Type Theoretical Meaning Reliability of Instrument Reliability
Analysis
Consistency of the measure of variables.
Cronbach’s Alpha is computed for each variable. It is commonly accepted that value above 0.7 are considered
acceptable (Nunnally, 1978)
Method of Data Analysis
Upon the completion of empirical data collection, the study will perform several of the various data analysis in best understand the population for this study. The major data analysis can be categorized into the following: Descriptive Statistics, Partial Least Square (PLS), Bootstrapping, R-Square, Path coefficient, correlation tests and Multiple Regression Analysis, to understand the from the ground level to the top, in a vertical understanding setting, from demographic distribution, in understanding the proportion of respondents, to the significant level of the research model, how is the relationship being formed, with a estimation of once the sample being duplicated, what are the possible pictures for the relationship, direction of effect as well as its significance with understanding of its correlation, and multiple regression result.
Descriptive Statistics
The Statistical Package for the Social Sciences (SPSS) 22 were used for the study, the setting is first use the SPSS in analyzing the empirical data collected from the respondents for descriptive statistics purposes, to learn about the distribution of the respondents such as distribution of the demographic, face validity and reliability test, the correlation table, as well as the distribution of the answers from respondents for each questionnaire items. All of which, the SPSS is a perfect tool for an analysis.
Further, the advanced analysis for an entire research model will be tested by the Partial Least Square, as it has some unique features, such as multivariate analysis, the analysis also allows researcher to replicate the sample, to examine a series of relationships, which is more suitable for the needs of this research model.
Partial Least Square (PLS)
Partial Least Square (PLS) is a type of structural equation modeling (SEM) technique described as a second generation of a multivariate analysis (Fornell et al., 1990). Instead of SEM technique, PLS allows researchers to replicate and examine a series of relationships. In terms of hypothetical constructs that are difficult to be measured theoretically, the estimation of PLS has no assumptions of normality and independence from observation.
Originated from the path analysis, PLS is utilized to work with small samples and exploratory studies. It has been regarded as an authoritative method on the theories of social science; the writer in the research adopts Smart PLS 3.2 as the major tool to scrutinize the relationships of hypotheses within all the variables. The concept of PLS is an iterative combination of component analysis relating measures to constructs. The significance of relationships among the constructs is measured as an indicator of the model performance. Therefore, by using PLS the techniques of bootstrapping, R-square, path coefficients and t-value are implemented to examine the structural model.
Bootstrapping
Bootstrapping is a way to duplicate the sample in a logical manner to retrieve the t-value for simulate the test with larger sample, to observe when sample size are expanded, would the result become significant. The bootstrapping assumes that sample is reasonably repeated of the assumed population distribution, and therefore the estimated coefficient in the PLS-SEM can be tested for their significance (Henseler, Ringle, and Sinkovics, 2009). The PLS algorithm uses the bootstrapping samples to estimate the result, such as the t-value and the path coefficient. Hair, Sarstedt, Ringle, and Mena (2012) also stated that critical t-values for the two-tailed test are 1.65, 1.96 and 2.58, which represented weak or one star, moderate or two stars, and strong relationship which represent three stars in significance to assess the relationship status.
Path Coefficient
The Path coefficients are used to examine, as well as to find the relationship between variables, particularly, to see if there’s significant factor identified, to understand its effect, the use of path coefficient provide a director of the effect, as well as its significance by identified the t-value, to present the significance in statistics or common research field, the denotation are usually expressed by significance at one, two or three star’s level, such way is a relatively simple and clear way to distinguish its significance, which the star are the higher the better, with respect value of 1.65, 1.95 and 2.56.
All Construct’s Illustration
In the following Table 3.3, is the illustration of all the constructs in terms of method of measurement for data analysis, each different constructs has its best assigned questions as well as measurement to best determine the effect of the construct, in order to provide a more realistic picture of the question as well as its weight given to the questions, this study use the table to illustrate its constructs, with number of questions assigned, as well as its measure, to have a glance look of the method of measurement in a relatively simple and effective way to understand.
The table will identified number of questions in each construct, the measurement which assigned to indicate the respondent’s answer, in best assess and helping researcher to distinguish the differences in respondent’s preferences, as well as, to understand the current status of each construct, to provide the best possible assessment for each of the different questions under various constructs.
Table 3.3. Constructs and Measurement of Data Analysis Constructs and Measurement of Data Analysis
Constructs N. of Questions Measurement Knowledge
Knowledge Enabler 20 1 = Strong Disagree
2 = Disagree 3 = Neutral 4 = Agree
5 = Strongly Agree Intellectual Capital 20 1 = Strong Disagree
2 = Disagree
Job Positions 1 1 = Board/Board Director
2= Senior Manager/Management