The research framework, hypotheses, sample, data collection, measurement and questionnaire design were be introduced in this chapter.
Research Framework
In this study, the independent variable is locus of control. The dependent variable is employee silence. Agreeableness and extraversion serve as the moderator on the relationship between locus of control and employee silence. The research framework is shown in Figure3.1.
Figure3.1 Research Framework
Research Hypotheses
The hypotheses of this study are proposed as followed:
Hypothesis 1. Locus of control is positively related to employee silence.
Hypothesis 2. Agreeableness will moderate the relationship between locus of control and employee silence. Under high level of Agreeableness, the positive relationship between locus of control and employee silence will be stronger.
Hypothesis 3. Extraversion will moderate the relationship between locus of control and employee silence. Under high level of extraversion, the positive relationship between locus of control and employee silence will be weaker.
Research Sample
Sample of this study were employees from nonprofit organizations. This study mainly focus on employees working in KMT and DPP Party. Convenient and snowball sampling were
H 2.
H 3. H 1.
Locus of control Employee silence
Agreeableness
Extraversion
20
used to reach more participants. A total of 302 participants filled the self-report questionnaire and there are 260 responses meeting the criterions.
The researcher choose political party employees to be sample because of following reasons. Since creating and maintaining a shared organizational value and culture is important for the operation of political parties (Jaskyte, 2004), employees’ communications with their supervisors and colleagues are thus essential. Therefore, examining determinant factors of employee silence behavior in political parties is very important.
Research Design
This study adopts a quantitative research design in order to examine the correlations among testing variables. Through questionnaire instruments, the quantitative data was collected. This study was used inferential statistics to test the hypotheses after the accomplishment of the data collection.
Questionnaire Design
The questionnaire comprises four parts measuring four variables of research interest:
locus of control, employee silence, agreeableness and extraversion. The questions used to measure all variables were derived from previous published studies in the literature. The survey contains five major sections. Section one contains six questions measuring locus of control, which was adopted from Lumpkins (1985) brief version of Rotters (1966).Section two contains seven questions testing employee silence from Vakola and Bouradas (2005). Section three contains ten questions testing extraversion. Section four contains ten questions testing agreeableness. Section five deals with respondents’ demographic information such as gender, age, educational backgrounds …, etc.
Since our target samples were non-profit organization employees in Taiwan, all the questions were translated into Chinese, and then translated back into English (back translation) by 2 professors who are fluent in both languages.
Instrumentation
The measurements used in this study were introduced in this section. The full questionnaire with instructions and measurements are presented in the Appendix:
Measurements and Questionnaire.
Employee Silence
21
The measurement items were from the scale developed by Vakola and Bouradas (2005).
The measure of employee silence contained seven items. An example item is: “how often do you express your disagreement regarding companies’ issues”. The Cronbach’s alpha was 0.87.
Responses were designed on a five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). Higher score indicates a greater degree of employee silence behavior.
Locus of Control
The measurement of locus of control was adopted the Lumpkins (1985) brief version of Rotters (1966). According to Lumpkin’s research, the Cronbach’s α coefficient of locus of control is 0.68.The measurement of locus of control contained six items. It included two dimensions, internal loci of control and external loci of control.
Agreeableness
The scale of agreeableness items were derived from IPIP (International Personality Item Pool; Goldberg, 2001) which was used to evaluate subjects’ five personality dimensions:
extraversion, agreeableness, conscientiousness, neuroticism, and openness to experience.
Each personality construct included 10 items (five worded positively and another five worded negatively). The measurement of agreeableness contained 10 items. There are four reversed questions. The higher score indicated that the person considered himself or herself has greater degree of agreeableness. The Cronbach’s alpha of this scale was 0.82. Responses were designed on a five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree).
Extraversion
The scale of extraversion items were also derived from Goldberg (2001).The measurement of extraversion contained 10 items .There are five reversed questions. The higher score indicated that the person considered himself or herself has greater degree of extraversion. The Cronbach’s alpha of this scale was 0.87. Responses were designed on a five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree).
Control Variable
Past studies have proposed that employee silence are influenced by gender (Brinsfield, 2009). Therefore, in order to minimize the influence of common method variance and clarify the correlations among locus of control, employee silence, agreeableness and extraversion, this study will control this variable.
22
Gilligan (1982) suggested that women's behavior are more relationship-oriented, therefore, when women voice their opinions they usually need to take other people's social relations and social acceptance into consideration. Therefore, compared to male, female show more silence behavior. Based on the study which has been published before, this study regards gender as control variable.
Pilot Test
The questionnaire was mainly composed by LOC scale, employee silence behavior scale and the big five personality trait scale .When the data was collected, the researcher conducted a pilot test to examine the internal reliability.
There are 40 participants from Democratic Progressive Party (DPP) completing the questionnaire and their age ranging from 24 to 58. Among them, there were 22 female and 18 male, 22 with bachelor degree and 18 with master degree in Taiwan.
After data was collected from these 40 participants, a reliability test was done using SPSS.
The Cronbach alpha value of each variable were 0.58 for LOC, 0.92 for employee silence, 0.87 for agreeableness and 0.82 for extraversion.
According to George and Mallery (2003), Cronbach Alpha value represented for the internal consistency of the item, ranging from 0.6 to 0.7 is acceptable, from 0.7 to 0.8 is good;
furthermore, over 0.9 is excellent. As a result, the overall measurement adopted in this study had a good reliability according to the results that pilot test indicated.
Data Collection
Snowball and convenient sampling approach were used for data collection. Participants received an e-mail with the link directing to the webpage of the survey while also encouraging them forwarding the email to as many of their friends as possible. Data collection period was lasted three months from July to August in 2017.After all the data was collected, the completed responses were analyzed through the following methods discussed in the next part.
Data Analysis
IBM SPSS 22.0 was used to perform data analysis. Confirmatory factor analysis (CFA) was used to test the reliability and validity of our measurement. Simple liner regression was performed to test hypothesis 1 and two individual hierarchical regression analysis were conducted to test hypothesis 2 and 3.
23
Descriptive Analysis
Descriptive statistics can help people have the whole images of the collected data. The researcher will use frequency distribution which including numbers and percentage to show the demographic information in this study, such as gender, age, tenure, education level and position level. The mean and standard deviation were utilized to investigate all the variables in this study, including locus of control, employee silence, agreeableness, extraversion and control variables.
Pearson Correlation Analysis
The liner relation between two continuous variables could be descripted as correlation, and researches adopted coefficient of correlation, also called Person’s product-moment correlation coefficient (Person’s r) to display the degree of the correlation. When the absolute value of Person’s r is closer to 1, the degree of the correlation is stronger. Also, the Person’s r is positive means the correlation is positive; on the contrary, the Person’s r is negative means the correlation is negative. In this study, correlation analysis was used to examine the relationship between control variables and locus of control, employee silence, agreeableness, and extraversion.
Hierarchical Regression Analysis
Hierarchical regression analysis was used to examine that does the interaction of locus of control, agreeableness and extraversion have influence on employee silence, in order to test the hypotheses 2 and 3 of this test. Control variables were put at the first model, and then the locus of control, agreeableness and extraversion were at the second model. Lastly, the interaction term of locus of control, agreeableness and extraversion were at the third model.
Research Procedure Determine the Research Topic
After discussions about the research topic with my thesis advisor for several times and reviewing the literature at the same time, a research topic mainly focus on political party employee’s silence behavior in KMT and DPP Party was determined.
Review the Literature
The researcher continuously reading the articles related to the variables, employee silence, locus of control, agreeableness and extraversion to get more background knowledge.
24
Finalize the Framework and Instruments
After determined the thesis framework, the researcher kept finding appropriate scale for every variables. The reliability and validity of each measurement were examined as well. At last, the questionnaires for each variable were chosen after discussing with thesis advisor；any revision for this research was adjusted accordingly.
Establish the Research Background, Purpose, Questions and Hypotheses
Researchers have discussed employee silence behavior in recent five years, which it is relatively a new research topic .Furthermore, the researchers mainly focusing on employees working in political party is quite rare in Taiwan. After reviewing the related literatures, the research background, purpose, questions and hypotheses were developed.
Data Collection
The sample of this study was chosen by convenience sampling and snowball sampling.
An explicit instruction and research purpose were demonstrated in the on-line questionnaire, and the scale was sent though the Internet. After 2 months, the data collection process was done.
Conduct the Pilot Test
In order to check the internal reliability of the data, conducting a pilot test is a must.
Analyze the Data and Discuss the Findings
After data was collected, researcher used statistical software including SPSS and M plus to analyze the data. Descriptive analysis, Pearson correlation analysis and hierarchical analysis were conducted in this research.
25