4. RESEARCH METHODS
4.1 Sample and Data Collection Procedure
The population of this study is the corporate governance awarded firms in Thailand at the level of good to excellent that are fully reported in the Corporate Governance Report of Thai Listed Companies (CGR) 2011. The CGR 2011 was presented in the database of the Securities and Exchange Commission on its website: http://www.sec.or.th. Based on this database, there are 361 companies as of February 1, 2011. In this study, accounting executive of each firm is the key informant.
The questionnaires are directly distributed to 361 companies; the valid mailing is 361surveys from which 83 responses are returned and usable. The effective response rate was approximately 22.99%. According to Aaker, Kumar and Day (2001) explained that the response rate for mail survey, with an approximately appropriate follow-up procedure, if greater than 20% is considered acceptable. Moreover, to protect possible response bias problems between respondents and non-respondents, a t- test comparison of the means of all variables between early and late respondents are tested corresponding with the test for non-response bias by Armstrong and Overton (1977). The results find no statistically significant difference between early and late respondents demonstrates represent non-respondents. Therefore, non response bias is not a problem in this study.
4.2 Questionnaire Development and Variable measurements
4.2.1 Questionnaire Development
In this study, a questionnaire to measure all constructs with multiple-item scales consists of seven parts. Part one asks for the personal information of key informant such as gender, age, marital status, education level, and experience. Part two contains the questions about the general information and history of business such as industry type, CG scoring, firm age, and number of employees. Part three through six are related to evaluating each of constructs in the conceptual model. These parts are designed by a five point Likert scale, ranging from 1(strongly disagree) to 5 (strongly agree). And part seven is an open-ended question for suggestions and opinions.
4.2.2 Variable Measurements
To measure each construct in the conceptual model, all of variables are gained from the survey and are measured by a five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). Additionally, constructs in the conceptual model are developed as new scales and modified from prior research. Therefore, the variable measurements of dependent variable, independent variables, moderating variables and control variables of this study are described as follows:
4. RESEARCH METHODS
4.1 Sample and Data Collection Procedure
The population of this study is the corporate governance awarded firms in Thailand at the level of good to excellent that are fully reported in the Corporate Governance Report of Thai Listed Companies (CGR) 2011. The CGR 2011 was presented in the database of the Securities and Exchange Commission on its website: http://www.sec.or.th. Based on this database, there are 361 companies as of February 1, 2011. In this study, accounting executive of each firm is the key informant.
The questionnaires are directly distributed to 361 companies; the valid mailing is 361surveys from which 83 responses are returned and usable. The effective response rate was approximately 22.99%. According to Aaker, Kumar and Day (2001) explained that the response rate for mail survey, with an approximately appropriate follow-up procedure, if greater than 20% is considered acceptable. Moreover, to protect possible response bias problems between respondents and non-respondents, a t- test comparison of the means of all variables between early and late respondents are tested corresponding with the test for non-response bias by Armstrong and Overton (1977). The results find no statistically significant difference between early and late respondents demonstrates represent non-respondents. Therefore, non response bias is not a problem in this study.
4.2 Questionnaire Development and Variable measurements
4.2.1 Questionnaire Development
In this study, a questionnaire to measure all constructs with multiple-item scales consists of seven parts. Part one asks for the personal information of key informant such as gender, age, marital status, education level, and experience. Part two contains the questions about the general information and history of business such as industry type, CG scoring, firm age, and number of employees. Part three through six are related to evaluating each of constructs in the conceptual model. These parts are designed by a five point Likert scale, ranging from 1(strongly disagree) to 5 (strongly agree). And part seven is an open-ended question for suggestions and opinions.
4.2.2 Variable Measurements
To measure each construct in the conceptual model, all of variables are gained from the survey and are measured by a five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). Additionally, constructs in the conceptual model are developed as new scales and modified from prior research. Therefore, the variable measurements of dependent variable, independent variables, moderating variables and control variables of this study are described as follows:
การแปล กรุณารอสักครู่..