This chapter covers research framework, hypothesis, sample and data collection, procedure and instruments. The hypotheses were tested through the measurement of OC, OCB, and KS on OI as demonstrated in figure 3.1.
Research Framework
According to the purpose of the research and literature review, figure 3.1 demonstrates the direct and indirect relationships among the organizational variables.
The relationships among the independent, dependent and mediating variables.
Figure 3.1. Research framework
Research Hypotheses
According to the literature review and the purpose of the study, the main and sub-hypotheses were formulated as follows.
Research Samples
A quantitative research method is employed for this study. The population for the study comprised of employees of different agencies within the public sector of Papua New Guinea such as central government, state government, local government and state-owned enterprises (SOE). These agencies include Department of Attorney and Justice, Department of Finance, Department of Treasury, Department of Personnel Management, Department of Labor, Internal Revenue Commission, PNG Customs Services, etc. This study used non-probability method of sampling in which snowball technique was employed to collect its data. In this technique, one or few research participants were recruited and these research participants then recruit other participants for the study in distributing and filling out the questionnaires. This technique was engaged due to the accessibility of members in each organization approached and the ease with which questionnaires can be distributed and data collected. Moreover, the technique enabled the researcher to reach a larger sample size resulting in a greater amount of data to work with.
The data was obtained from the individual respondents who answered research questionnaires supplied. These questionnaires were distributed to employees of all the
agencies within the public sector in Papua New Guinea and took up to a maximum of two months for completion. The questionnaires were provided with both online surveys and pen-and-paper surveys format. The total participants consisted of 301 employees with demographic profile characteristics under investigation inclusive of gender, age, education level and position. Respondents remained completely anonymous. In addition, to encourage respondents’ participation, the employers’ assistance was sought to allow employees to answer the survey during office hours.
Research Procedure
This section explains the procedures engaged in this research. The method employed in this study were divided into various stages and presented in the figure 3.2.
Each stage described each action taken during the research. The literature review was done in the initial stage and the research topic was identified and decided. Then research questions were formulated according the literature review. Based on the literature review, research framework was also developed followed by research hypothesis.
Subsequently, research instruments were identified and minor adjustments were performed to ensure easiness in understanding, rationality in consistencies, and contextual relevance. Sample selection was identified and questionnaires were distributed and data were collected. After data was collected, it was analyzed. At the final stage, the findings of the study were generated and discussion and conclusion were produced.
Figure 3.2. Summary of research procedure
Research Instruments
This study used measurement items obtained from the literature and to warrant contextual consistency, minor adjustments were made to the items. Each question contains the four variables; OC, OCB, KS and OI, and each respondent completed the questions for the four variables. Each variable to measure the constructs are labelled with its Cronbach’s alpha. Items of the OCB (a =.93) were adopted from Podsakoff, (2000). KS items were adopted from Bock, Zmud, Kim and Lee (2005). The items of KS (a =.89). Items of OC (a =.87), were adopted from Allen and Mayer (1990). Items of OI include these three; 1. Process Innovation (a = .85), 2. Product Innovation (a = .81), and 3. Administrative Innovation (a = .92) and were obtained from Miller and Friesen’s (1983). The items used in measuring the constructs are described as the following: Organizational commitment (OC) has 7 items, organizational citizenship behavior (OCB) has 12 items, knowledge sharing has 10 items and organizational innovation has 9 items. The scales of the items were measured on a 5-point Likert Scale ranging from strongly disagree (1) to strongly agree (5). Towards the end of the survey, several questions were asked to obtain demographic information of the respondents such as gender, age, position and education level.
A pretest was carried out to make sure that the reliability and face validity of the questions are acceptable according to standard. Three professors were engaged who assessed the items in terms of instrument clarity, question wording, ease of understanding, logical consistencies, sequence of items, and contextual relevance.
Their feedbacks were used to revise problematic items.
Data Analysis
Statistical Package for the Social Science (SPSS) software was used to analyze the collected data. The final items used to complete the data collection were 38. Each item is coded as given in the table 3.1.
Table 3.1.
Reliability and Validity
In quantitative research, experimental methods and quantitative measures are used to test the hypothesis (Hoepfl, 1997), and the measurement and analysis of causal relationships between variables are also emphasized (Denzin & Lincoln, 1998). All measurements, particularly measurements of behaviors, opinions, and constructs, are subject to fluctuations or error that can affect the measurement’s reliability and validity.
Reliability measure the consistency of items to ensure its accuracy in representing the total population under study while validity determines whether the research truly measures that which it intended to measure or how truthful the research results are (Joppe, 2000). The reliability was measured using Cronbach’s alpha (internal consistency) and it was computed for each variable. The value of the Cronbach’s alpha was 0.7 or above to be considered acceptable.
Descriptive Statistics
Descriptive statistics are used to describe the basic features of the data in a study. It provides simple summaries about the sample and the measures. However, it should not be used to make conclusions beyond the data that have been analyzed or reach conclusions regarding any hypotheses that have been made. Descriptive statistics are very important in explaining raw data because raw data cannot be simply presented or else it would be hard to visualize what the data is showing, especially if there is a lot of it. Descriptive statistics therefore enables the research to present the data in a more meaningful way, which allows simpler interpretation of the data. Usually data are described by two types of statistic which include:
Measures of central tendency: It refers to the approach used in describing the central position of a frequency distribution for a group of data.
Measures of spread: it refers to approach used in summarizing a group of data by
describing how spread out the data is. The method is used to assist in interpreting the data by calculating the mean, range, quartile, variances and standard deviation.
Correlation and Regression
To calculate the variation and strength of relationship between two variables the method of Correlation analysis was applied. This relationship was expressed in the form of an equation as the Regression. To further test for the probability of a linear relationship between the variables a correlation coefficient, like the Pearson Product Moment Correlation Coefficient, was used. The calculation of this correlation coefficient (r) would lead to the quantification of the strength of the relationship. A numerical value ranging from +1.0 to -1.0. r > 0 indicated positive linear relationship, a range where r < 0 indicated a negative linear relationship while r = 0 indicated no linear relationship. In regression analysis, it is the nature of the relationship itself between the dependent variable (response) and the independent variable (explanatory) that presents the problem of interest. To test the mediating effect of knowledge sharing as a variable hierarchical regression was performed. The analysis consists of choosing and fitting an appropriate model, done by the method of least squares, with a view to exploiting the relationship between the variables to help estimate the expected response for a given value of the independent variable.