A study of the factors contributing to the success of industrial cluster implementation in Thailand

Abstract

This research studies the factors contributing to the success of industrial cluster implementation in Thailand. The objectives of this research are: 1) to study the criteria of effectiveness and efficiency, 2) to identify the factors that lead to successful industrial cluster implementation in Thailand from an implementation perspective, and 3) to receive suggestions for improving the implementation of the industrial clusters. The research is conducted at the enterprise level of analysis. Secondary data from questionnaires belonging to the Department of Industrial Promotion (DIP) were used for the research. The respondents were enterprises that joined the industrial cluster project implemented by the DIP in fiscal year 2007. During that year, the DIP implemented 22 industrial clusters in various areas in Thailand. The questionnaires that were responded to were collected randomly for each industrial cluster with an expectation of 10 answered questionnaires per industrial cluster. Therefore, the questionnaires represented 220 enterprises. A Likert scale was used for the questionnaires but four kinds of Likert scales appeared in the research: 1) a 10-point Likert scale from level 1 to level 10; 2) a 5-point Likert scale from level 0 to level 5 and level 1 to level 5; 3) a 3-point Likert scale from level 1 to level 3; and 4) the decision to say “yes” or “no.” The different levels ranged from strongly disagree at level 1 to higher levels of agreement; level 0 meant “no opinion.” For the research methodology, cross-tabulation was used to study such characteristics of fundamental statistics as frequency distribution and percentage. The disaggregate method was used to disaggregate the data that were collected from the cluster level of analysis to the enterprise level of analysis; this method distributed the data from 22 units to 220 units. Replacing the mission value by series mean was used as a way to replace the missing value by approximation. A z-score was used in the case of different constructing scales, this being one of the methods of standardizing data. Structural equation modeling (SEM) was used for simultaneous study of a series of interrelate dependence relationships among variables or indicators. Overall, the findings of the study provided interpretations, suggestions, limitations, and recommendations for future research. The findings also served to draw implications for research, practice, and theory.