ISSSM-0267 A Study on the Performances of Geometrical Habits of Mind of Primary School Children Chia-Huang Chen Department of Mathematics Education, National Taichung University of Education, Taiwan E-mail: [email protected] 1. Background Geometry is the field in which students explore and develop mathematical concepts. The ability to develop geometric reasoning in graphics and space can support students' mathematical habits of mind and provide a substantial guarantee of knowledge of output. Goldenberg and Cuoco (1998) argue that geometry can help students connect mathematics-related concepts and abilities, develop students' generalized, inferential habits of mind, and provide different case designs that allow students to demonstrate the claims of graphics and space objects (Ellis, 2011), in the process of solving problems or research, understand how mathematicians deal with mathematics. The layout of geometric textbooks is related to the acquisition and development of students' cognitive ability and problem-solving ability. The perfect teaching can help students to establish geometric concepts, learn concepts from calculation, and solve geometric problems with geometric knowledge. Therefore, the development and cultivation of geometric habits of mind is an important issue and worthwhile exploration. Connecting habits of mind means having an attitude toward intellectual formation in the face of problems that cannot immediately explain the problem-solving method (Costa, 2000). Costa (2000) believes that students' connections to these intelligences are fundamental, and that they can only become attitudes by examining and acting on the behaviors suggested by the habits of mind. The concept of graphics and space provides important content for students' attention, performance, argumentation, definition and relationship reasoning (Cuoco et al., 2010; Kobiela & Lehrer, 2015). In view of the importance of mathematics and geometric habits of mind, these sixth grades as this study’s sample, and for graphic and spatial tasks, it is required to describe the problem clues and solve the questions and collect interviews, analyze the data, and further understand the students. The performance of geometric habits of mind provides practical application in the future. The specific inquiry questions are: 1. What are the performances of the habits of mind of the sixth grades in graphic and spatial issues by the study? 2. What are the ways in which the students think about solving problems in the graphic and spatial problems provided? 398
3. What are the key features of the performance of the mathematics and habits of mind of students who have succeeded or failed in solving problems in graphics and space? 2. Methods Research Sample The participants in this study were from a total of 105 students in the sixth grade of five public elementary schools in central Taiwan. The study was conducted from May to June 2017. These sixth-grade students have studied the basic geometrical experience of the perimeter and area, polygons, cylinders and vertebral bodies, and application formula solving. The Task Design In order to test whether sixth-grade students have the relationship of exploring geometry and the reasoning of these relationships, the concept of generalized geometry, exploring the variables and invariance of these relationships, and using these elements to draw geometric images and other mental habits, this study designed three graphics and space tasks, including “sector” “isosceles trapezoid” and “cuboid”. Data Collection and Analysis The data is divided into two parts for collection: one is to ask students to perform the performance of drawing and calculation according to the provided clues, and the researchers classify according to the characteristics of the students' performance, the types include: 1). No answer, no performance content; 2). Wrong response, drawing or calculation is incomplete, wrong; 3). Drawing is correct but the calculation is wrong, according to the provided task clues correctly draw the graphic or spatial image, but the requirements Geometric object relationship calculation error; 4). Correct calculation but drawing error, according to the provided task clues, the required geometric object relationship calculation error, but draw the graphic or spatial image; 5). The calculation and drawing are correct, etc. 5 kinds of performance. And a percentage statistical description of the number of performers of the type. The other data is classified according to the student's thinking path. The data analysis firstly analyzes the percentage of the thinking path of all the students, and then tries to make the correct drawing but the calculation is wrong, the calculation is correct but the drawing is wrong, the calculation and drawing are correct, etc. Statistical analysis of the percentage of thinking paths of performance. In addition, interviews were conducted. A total of 12 interview samples were selected based on the students' correct performance and wrong performance in the three tasks. Task 1 and Task 2 select one from each of the correct performers and two from the wrong performers. Task 3 is based on the correct number of people, so 2 are selected, and 4 wrong performers are selected. In 20 minutes, the content of the interview was explained by the erroneous problem solving and the performance of the correct problem solver to present the geometrical habits of the graphical and 399
spatial tasks. Student graphics and space tasks are solved by individual paper and pencil test methods. Answers are made according to the task sequence. You can't jump and answer. Each task completes the drawing and calculation, and then sorts the path of solving the problem. The total time for the three tasks is 40 minutes. 3. Results This study designed three graphical and spatial tasks to explore the students' geometrical habits of mind and graphic and spatial concepts. The subjects were from 105 students in the sixth grade of five public primary schools in central Taiwan. According to the clues provided, students are required to perform the performance of drawing and calculation, collect student data in accordance with the interview method, and describe the research results in qualitative analysis and quantitative way. The research find: The sixth-grade students who participated in the research provided poor geometrical habits of mind. Most of the students had poor geometric mental performance. They could not explore the relationship between the geometric elements of the sector, isosceles trapezoid and the cuboid and the reasoning and generalization of these relationships. Concepts, explore the variables and invariances in these relationships, use these elements to evaluate and draw geometric images, and use their features to solve problems. Students' problem solving for graphic and space tasks could roughly divided into two types of thinking: the overall image and the elemental analysis. The accuracy of using the former to solve the problem is higher than that of the latter. The correctness of the problem of graphic and spatial tasks is different from that of the wrong person. The correct problem solver can clearly understand the relationship between the elements that constitute the geometric object correlation, and can use them to further reason and link to draw the correct geometric figure. Like and use the generalized formula to solve the problem, and those who solve the wrong problem think about using the formula to solve the problem, and do not think and reason the relationship between the geometric elements, so that the problem is wrong. Keywords: Geometric, Habits of mind, Graphic and spatial, Mathematical literacy. 400
4. References Costa, A. L. (Ed.) (2000). Developing minds: A resource book for teaching thinking. Alexandria, VA: Association for Supervision and Curriculum Development. Cuoco, A., Goldenberg, E. P., & Mark, J. (2010). Contemporary curriculum issues: Organizing a curriculum around mathematical habits of mind. Mathematics Teacher, 103(9), 682-688. Ellis, A. B. (2011). Generalizing-promoting actions: How classroom collaborations can support students’ mathematical generalizations. Journal for Research in Mathematics Education. 42(4), 308–345. Goldenberg, E. P., & Cuoco, A. A. (1998). What is dynamic geometry? In R. Lehrer & D. Chazan (Eds.), Designing learning environments for developing understanding of geometry and space (pp. 351–367). Mahwah, NJ: Erlbaum. Kobiela, M., & Lehrer, R. (2015). The codevelopment of mathematical concepts and the practice of defining. Journal for research in Mathematics Education, 46(4), 423-454. 401
ISSSM-0289 Course Timetabling and Classroom Arrangement Planning Model Chun-Ying Chena, Yi-Fan Jiangb Department of Air Transportation, Kainan University, Taiwan E-mail: [email protected] a, [email protected] 1. Background The problem of course timetabling and classroom arrangement is a routine matter that every school administrator must do every semester. It is also a very important part of the school's administrative work. However, due to the limitations and complexity of scheduling problems, it is more to optimize the scheduling. In general, it is necessary to consider the number of classrooms, the capacity of classroom, the number of teachers, the time of the lectures, the students' rushing halls, the equipment resources, the hour limit, and so on. The course timetabling and classroom arrangement problem is cumbersome in schools. In order to satisfy the various conditions of the class scheduling problem as much as possible, most of the practice is still manual, relying on the experience accumulated by the administrative staff for a long time to satisfy these complex constraints. However, in fact, the preliminary results of the class arrangement are not satified all the restrictions. It is necessary to repeatedly modify and communicate with the teachers. Usually, under the constraints of the time, only a feasible solution can be found. But in fact, it is not really to find the optimal solution, and the results of the last class schedule are not necessarily in line with everyone's expectations. This kind of practice is quite time-consuming and human resources. The output schedule often depends on the subjective consciousness of the class staff. It is not only lacking in objectivity, but also difficult to measure the quality of the output schedule. It is quite lack of efficiency and effectiveness. In addition, the current education system is very different from the past. For example, many new courses have been introduced, so that the degree of freedom of students' elective courses is more extensive, and the school pays more attention to the requirements and quality of teaching. Therefore, the purpose of class scheduling should be to consider how to meet the needs of students, and at the same time take into account the quality of teaching that teachers hope to achieve. These soft limits will undoubtedly make the final solution set very large, increase the difficulty of solving the algorithm and increase the space for narrowing the solution, and even cause the situation without feasible solution. This study will establish a set of integrated mathematical planning models for the course units, each unit can adjust the required restrictions according to their needs, and finally confirm the model and proposed solutions used in this study by examples. 2. Methods 402
The input data for the optimization of the course timetabling and classroom arrangement problem can be divided into two parts: the teacher and the course. The teacher's information includes the time that the teacher can arrange the class and the course taught by the teacher. The course materials include classrooms for each course and time slots for each course. The output information can be divided into (1) the time of each teacher's class, (2) Each teacher teaches course, and (3) Classrooms used in each course.We ues mathematical programming to develop an optimal course timetabling and classroom arrangement model. The objective function (1) is to minimize the time slot of course that students do not expect. For example, the first lesson every day, Friday afternoon class. Constraint (2) ensures that each teacher can only have one class at a time. Constraint (3) ensures that each classroom can only arrange up to one class at a time. Constraint (4) ensures that classes 6 and 7 may not be scheduled on Wednesdays (School requirements). Constraint (5) ensures that some courses may not be restricted in certain classrooms (including computer classrooms and number of people). Constraint (6) ensures that the class of day division, continuing education division and MBA classes are arranged to the suitable time slpt. not subject to time limit. Constraint (7) ensures that it is satisfied teacher's time limit for teaching. Constraint (8) The same year of the same grade course (including mandatory, elective) can not be rushed. Constraint (9) avoids the teacher's idle time being too long and avoids the teacher's schedule on the same day in the first and eleventh sections. 3. Results This study is solved the course timetabling and classroom arrangement planning problem. Through a large number of literature reviews and current situation analysis, we understand the output and input data of this research question, and then through the writing of the computer program, solve the problem of course timetabling and classroom arrangement. Finally, the test was conducted to examine the feasibility and correctness of the study. The schedule helps reduce the workload of administrative staff. Keywords: Mathematical programming, course timetabling, classroom arrangement 4. References Abdelhalim, E. A., and Khayat, G. A. (2016). An information visibility-based university timetabling for efficient use of learning spaces (IVUT). Egyptian Informatics Journa, 17(3), 315-325. Babaei, H., Karimpour, J. and Hadidi, A. (2015). A survey of approaches for university course timetabling problem. Computers & Industrial Engineering, 86, 43-59. Carter, M. W. and Laporte, G. (1997). Recent developments in practical course timetabling. International conference on the practice and theory of automated timetabling, pp.3-19. 403
ISSSM-0294 A Micropolitical Analysis of Implementation of the Multi-Grade Instruction in Remote and Small Schools—Using a Public Elementary School in Central Taiwan as an Example Ya Ching Chenga, Ya Feng Chengb aChanghua County Tuku Elementary School bUniversity of Kang Ning E-mail: [email protected] 1. Background With the declining birthrate and population concentration in cities, the number of elementary school students in rural Taiwan has sharply decreased, resulting in a high turnover rate of teachers and school principals. To avoid closure of schools, principals are seeking ways to improve the teaching quality of their teachers and the multi-grade instruction is one of solutions. This study employed the micropolitical analysis method to investigate the exercise of political power in the implementation of the multi-grade instruction and its influence on teaching quality. 2. Method Located in a plain area with thirty-seven students in Central Taiwan, this elementary school has implemented a multi-grade teaching program for two years, in hope of gaining more subsidies and improving the teaching quality. As an executor of this program, the author of this study collected qualitative data from Aug 2016 to Aug 2018 to analyze the implementation of this program, the intergroup competitions within this school, and the changes in teaching pattern through three different principals. Besides, relevant literature review was conducted and experts were invited to assist in validating analysis results. 3. Results (1) Schools must make changes Through implementation of the multi-grade instruction, the number of students for collaborative learning increases and the flexibility in teaching styles rises. Besides, students are able to receive considerable attention from teachers as in a small school and have opportunities to strengthen their competitiveness as in a large one. The saved teaching time can be utilized to support after-school care and club activities, alleviating the burden of funding. There is another factor that the school was assigned by the Head of Education Bureau to implement the program. The additional subsidy for this program is useful to the school. (2) The principal’s leadership style affects the speed of change The principal’s leadership style has an impact on the power structure of the teachers’ interest 404
group. The resulting campus culture affects the speed at which changes in instruction are carried out. In fact, most teachers oppose changes. Over a span of just two years, three principals with different leadership styles were involved. When the principal adopted a transactional leadership style, most teachers focus more on power struggles and less on teaching quality. When the principal adopted a centralized and reasoning leadership style, teachers tended to form alliances that run under the table, but teaching dialogue began to take place. When the principal adopted an empowering leadership style, teachers would be demanded to focus on teaching quality. (3) Look for breakthroughs and set up a system first Multi-grade instruction is a set policy. Amid resistance, isolation, and conflicts, the Director of Educational and Student Affairs Division and the Chief of Educational Affairs Section sought breakthroughs in three aspects. Firstly, to convince external teachers, substitute teachers and formal teachers to participate in the plan. Within two years, multi-grade instruction has been extensively applied in reading, physical education, arts and skills, general curriculum, student club, and remedial instructions. Secondly, to invite experts to offer guidance and consultation. The measures were intended to increase teachers’ understanding of the meaning and implementation method of the multi-grade instruction and to reduce their resistance. Thirdly, to request teachers to compile a syllabus and teaching plans. Their experiences in providing multi-grade instructions were gathered and could be shared. (4) Authoritative leadership and democratic communication working in parallel, and teaching effectiveness is the focus The principal is the key figure in the making of all school decisions. Each semester, a team from the university of education would visit the school to observe the program outcomes. All the teachers were aware of these facts. The Educational and Student Affairs Division has allowed teachers to express their opinions through the willingness questionnaires and the feedback sheets. The results would be quantified and submitted to the principal. The Director would also announce and explain the survey results in morning meetings, making “teaching effectiveness” a shared goal among all members of the school. Keywords: micropolitics, multi-grade, remote and small school References Chen, H. J. (2013). Examining intergroup conflict: The micropolitical analysis of a small school. Journal of National Pingtung University of Education: Education, 40, 243-272. Chen, H. J. & Yu, C. J. (2011). An analysis of the micropolitics in the curriculum development of school features in an elementary school. Educational Resources and Research Bimonthly, 103, 143-172. Chen, C. C. (2001). The micropolitics of distributed leadership. Journal of Education National Chunghua University of Education, 18, 1-28. 405
ISSSM-0330 The Reflexivity and Life Politics in Curriculum Policy Making Teng Huang Fu Jen Catholic University E-mail: [email protected] Abstract M. Apple’s curriculum theory is the key theory explaining the process of curriculum development and struggle. However, Apple’ s theory seems to have its limitation in the context of reflexive modernity. In this paper, I first present the three presumptions in Apple’s theory. The three presumptions were as follows. Firstly, in Apple’s theory, the national hegemony in national curriculum and textbook is transmitted from top to down. It starts from the struggles among different political groups. But when certain group wins in the end of struggles, they get the power to control educational system and infuses the ideologies to maintain their position. Secondly, related with the first point, the process of political struggle basically happened outside the field of education rather than inside the field. For example, in Apple’s later work, he highlights the groups of Neoliberalism and Neoconservativism as the main forces of the politics of educational reform, where the educational groups seem to have only little influence in the process. Thirdly, in the top-down process, educators, teachers and students are passive to accept the top-down social control and ideology dominated by hegemony. These stakeholders of educational systems have little room to resist the hegemony as if they are determined and have no agency to fight back. However, in this paper, I will argue that, through the real event of curriculum policy making happened in Taiwan year 2014, I argue that these three presumptions are no longer to explain the real situation in current society. Instead, drawing from the theory of reflexive modernity and life politics, I suggest that new concepts of curriculum policy making should be addressed and thus I propose some new perspectives to interpret the new context of curriculum policy making. Keywords: curriculum policy making, reflexive modernity, life politics 406
ISSSM-0384 The Analysis of the Educational Policy for Elementary and Secondary School Curriculum in South Korea SeonMee Kim Center for Global Multicultural Education, Sunchon National University, South Korea E-mail address: [email protected] 1. Background Since schools in South Korea have begun to use new curriculum from this year, it is important to understand the characteristics of the revised curriculum for its successful implementation in classroom teaching. This paper attempts to figure out the focus of new curriculum, and to examine the aspects of its appropriateness for the classroom practice in reality, and rapid changing social and technological environments. 2. Methods The main questions of this work are: what are the characteristics of new school curriculum for elementary and secondary education?, what changes are made for future education?, and whether the curriculum changes are suitable to meet with educational challenges for the future educational circumstances. This paper investigated and analyzed the school official curriculum documents revised in 2015, including the general curriculum guidelines book, elementary, middle, and high school curriculum documents. The government reports of education plan were referenced in order to understand the main policies of the future school curriculum innovation. Other recent research works conducted from 2010 through 2017 were also thoroughly examined to figure out ideas of future education and the characteristics of school curriculum. 3. Results The curriculum adopted from this year had been made in order to minimize the problems in previous educational practice, such as knowledge memorization instead of creative thinking, segmentation of each subject matter, and to be in accordance with the requirement of new competencies suitable for revolutionary social change nowadays(Kim, et. al., 2017). For short, the main goal of the new curriculum is ‘people with creative & integrated skills’, and six competencies are emphasized, that is: self-management, knowledge & information operating, creative thinking, aesthetic emotion, communication skill, and social skill. The characteristics of the curriculum are listed below(Ministry of education, 2015a). - To emphasize student’s centered curriculum and to increase student’s initiation, freedom, and the choice of curriculum contents and subject matters. - To reduce the amount of knowledge for study, and to improve the quality of learning experience. 407
- To extend student’s activity-based learning and participation, and process-based performance evaluation. - To diversify curriculum contents, and more freedom of curriculum management and practice in schools. More concretely, elementary school curriculum emphasizes the integrated curriculum, computational thinking & SW education, role-play methods and reading for liberal art education, and the education for safety(Ministry of education, 2015b), as the results of the disaster in Danwon high school. In middle school, a free semester or a free year system is introduced, which intends flexible curriculum operation emphasizing students’ discussion, learning by direct experiences, and participation, with no achievement tests during this period(Ministry of education, 2015c). A student’s self-managed learning process and the experience of vocational inquiries are also important factors. The notable changes in high school curriculum is that the division of liberal art and science tracts is removed, so as that students choose any subject of study and they are able to integrate any subject matters(Ministry of education, 2015d). The integrated social science textbook and integrated science textbook are introduced, elective subject matters are diversified. On the other hand, the vocational high school adopted the customized curriculum contents suggested by national curriculum standard(NCS). The digital textbooks are introduced in all school levels, yet in English, social studies, and science subject matters only. As Choi(2014) pointed out, schools still maintain the unit system for management of curriculum, considered unnecessary tradition. In the whole, the official school curriculum made documental changes in many aspects and teaching methods, yet it is suggested to require lots of works to reach at real change of classroom practice. Keywords: School curriculum, Elementary & secondary education, Educational policy 4. References Choi, J.(2014). Problems in adopting and applying of unit system for high-school curriculum in Korea. The Journal of educational research, 12(4), 43-69. South Korea. Kim, S., Bang, G., Kim, J., Seo, D., Woo, J., Ryu, E., Ryu, J., Kang, C., Sim, H.(2017). A study for the development of educational policies to implement 2015 revised curriculum in Jeonnam schools. Jeonranam-do, South Korea: Office of education. Ministry of Education(2015a) The 2015 Curriculum (main book). South Korea. Ministry of Education(2015b) The 2015 Curriculum (elementary school) South Korea. Ministry of Education(2015c) The 2015 Curriculum (middle school) South Korea . Ministry of Education(2015d) The 2015 Curriculum (high school) South Korea . 408
ISSSM-0266 Asymmetric Responses of News Coverage to the Stock Prices of Domestic and Offshore Companies Listed on the Taiwan Stock Exchange Ming-Chang Wang*, Chien-Chuan Ko, Fei-Han Lee Department of Business Administration, National Chung Cheng University, Taiwan * E-mail: [email protected] Abstract This study examines the asymmetric responses of the stock prices of domestic and overseas companies listed on the Taiwan Stock Exchange to news coverage. Conducting term training on the news coverage and stock prices of the constituent stocks of TSEC Taiwan 50 using the support vector machine algorithm, which serves as a reference to categorise positive and negative news coverage, this paper explores the responses of press coverage on the pricing behaviour between offshore companies and the control samples as domestic listed companies. Compared with domestic listed companies, we find that stock prices of offshore listed companies rise significantly on the day of positive news release, but when negative press coverage is released, the two groups of companies exhibit insignificant difference in stock price decline, representing that positive and negative news coverage leads to asymmetric responses to stock prices. In addition, our evidences show that when the volatility index climbs (declines) materially, positive press coverage leads to less (more) significant increases in the stock prices of offshore listed companies than those of domestic listed companies, but the stock prices of the domestic and offshore listed companies from negative press coverage releases both decline, indicating no significant difference, also leading to asymmetric responses to the stock prices of domestic and offshore companies listed in Taiwan. Keywords: information asymmetry, media management, investor sentiment, text mining, support vector machines I. Introduction Following the development trend of capital markets worldwide, the Taiwanese government launched the “1-2-3 Project for Exchange Listing by Offshore Firms” in 2008 to attract offshore companies to list on the domestic securities market. This has efficiently expedited the development of the capital market in Taiwan by providing incentives to foreign firms. As of March 2016, a total of 49 offshore companies have been approved to list in Taiwan, with their principal place of business in China, nearby Asian countries, and the United States. Among them, the majority of the offshore companies are registered in the British Cayman Islands. However, because the places of business, registration, and listing are in foreign countries, offshore companies listed in Taiwan encounter challenges in obtaining and disclosing information 409
simultaneously, which further causes low information accessibility and accuracy. Research by Brown and Hillegeist (2007) indicates that the quality of information disclosed can affect levels of information asymmetry, which highlights that higher disclosure quality can lower the level of information asymmetry. For overseas companies listed in Taiwan, because both their places of operation and registration are located overseas, investors tend to have difficulties obtaining their information, and that the information obtained typically has relatively low accuracy. Because positive and negative news coverage released by media platforms causes immediate and effective investment decision-making by investors, the information quality of news coverage for listed companies can affect stock price behaviour. When investors perceive the quality of news coverage for certain listed companies (e.g., offshore companies) to be inferior, they evaluate investment risks from a subjective perspective to adjust their investment decisions, resulting in direct effects on stock prices. Therefore, the quality of news coverage can simultaneously affect stock price behaviour. Moreover, extant studies have justified that investment distance can influence investors’ decision-making. Liao, Li, Zhang, and Zhu (2012) assert that investors opt to invest in shares listed on nearby exchanges, and furthermore, Pirinsky and Wong (2006) emphasise that investors tend to invest in listed companies whose headquarters are near their location. These extant research results highlight that when the distance between listed companies and investors is far, investors may have difficulty accessing the economic information of listed companies, which causes challenges in validating information accuracy. Consequently, greater challenges in evaluating such listed companies’ enterprise value results in a higher level of information asymmetry, thereby reducing investors’ willingness to own the stocks of such companies. Differences in the geographical distance and levels of market familiarity cause gaps between domestic and offshore listed companies in the quality of information they disclosed, thereby leading to dissimilar levels of market information asymmetry. Therefore, this study examines the responses generated by various levels of information asymmetry on the stock prices of the domestic and offshore companies listed on Taiwan’s securities market. In the capital market in Taiwan, offshore listed companies have a higher level of information asymmetry than domestic listed companies, which results from four primary characteristics. First, compared with domestic listed companies, the geographic distance of offshore listed companies is far for investors to access information. The locations of company registration, business offices, and operations are outside Taiwan’s territory, which generates restrictions such as time difference, geographical distance, and contents of business operations, thereby affecting the levels of information disclosure. From the perspective of domestic investors, the information of offshore companies is less accessible than that of domestic companies, which restricts investors from accessing sufficient information in a timely manner, thereby leading to a higher probability of information asymmetry and higher investment risks. Although the Taiwan Stock Exchange 410
(TWSE) strengthens the information disclosure of listed companies with its Market Observation Post System (MOPS), the volume, quality, and time-efficiency of nonfinancial information (e.g., news coverage) for offshore companies have remained relatively unfavourable for domestic investors. Therefore, the challenge of information validation significantly increases the level of information asymmetry in offshore companies, thereby indirectly affecting stock price behaviour. Second, offshore listed companies have mostly conducted reorganisation to comply with the listing rules of Taiwan. To comply with the legal requirements of primary listing, many offshore companies have changed their organisation structure through reconstruction or reorganisation (e.g., cross ownership and stock dilution) before submitting listing applications. However, multiple rounds of reorganisation often result in various problems, such as unclear internal and external messages and insufficient information disclosure. For example, the Japanese company Elpida (TWSE stock code: 916665) issued Taiwan depositary receipts (TDR) for the securities market in Taiwan in 2012. To comply with Japan’s Corporate Reorganisation Law, Elpida conducted fundamental financial and business reconstruction, which qualified the company to apply for the reorganisation process; specifically, they could apply for bankruptcy protection. This application led to a limit down of the TDR stock price by 98% in consecutive days, which unfortunately resulted in delisting on the TWSE. Although the investors claimed compensation, the reorganisation process was conducted in the Tokyo Stock Exchange. The handling of Elpida’s TDRs should have been equivalent to that of the stocks listed on the original country’s exchange, which made Taiwanese investors suffered great financial losses. Elpida was fined by TWSE for severely violating the listing rules for information disclosure. Specifically, it failed to inform Taiwanese investors timely about its trading suspension in the Tokyo Stock Exchange. However, the serious level of information asymmetry had deepened the losses and distrust of Taiwanese investors. Third, all offshore listed companies register their companies in tax havens. Several cases of fraud as well as financial crises have occurred in Taiwan’s securities market, which have urged the government to implement stricter and more conservative listing rules than other Asian countries. However, to overcome internal operational challenges and in consideration of external competitors, domestic enterprises have low willingness to disclose information to external parties, which increases the pressure of information disclosure. This problem pushes them to register companies in third countries; thus, the offshore companies owned by domestic enterprises can apply to list on Taiwan’s capital market. Currently, the majority of offshore companies listed in Taiwan are registered in the British Caymen Islands. Because their registration, business operation, and listing locations are in different countries, their operations and profits are easily affected by global political and economic situations and exchange rate fluctuations, which influences their revenue and performance. More crucially, from the perspective of domestic investors, listed companies register in tax havens to circumvent 411
regulations, evade taxes, and benefit their business, thereby intensifying the level of information asymmetry. Fourth, the government offers comparatively looser rules to attract offshore companies to list in Taiwan. Investing in offshore companies is riskier than domestic investment. To attract foreign companies to apply for primary listing in Taiwan, TWSE merely accepts listing applications from foreign companies that have never been listed in securities exchanges in any other countries, as well as offers loose rules (e.g., the criteria of company scale, profiting capability, and number of stocks for sales) to eligible foreign companies. Whereas foreign listed companies receive the incentive of business income tax exemption, domestic investors can be exempt from income tax for a certain amount of stock dividends. However, looser rules for offshore companies listing applications imply higher levels of information asymmetry, which facilities these companies to hide related operational information. Accordingly, the stock prices of offshore listed companies may be overestimated because of insufficient disclosure. To encourage foreign companies to invest in Taiwan, the government in Taiwan has improved the internationalisation of the capital market and actively developed the domestic market to serve as an emerging platform in Asia for funding and investment. A press release from the TWSE in December 2015 indicates that the performance of KY-stocks issued by offshore listed companies generated an average price-earnings ratio of 26, turnover rate of 182%, and yield rate of 3.4%. This outstanding performance not only encourages foreign institutional investors but also domestic investors to invest KY-stocks. However, the excellent performance of KY-stocks attracts vulture investors from abroad. In 2014, the international vulture institution Gluacus made a great deal short selling for stocks issued by an offshore company, Asia Plastic Recycling Holding Limited (stock code: F-1337), and then distributed stories to the press regarding financial statement fraud. This pushed down the stock prices of Asia Plastic for 11 consecutive days in the TWSE. The resulting high selling pressure in the market affected not only the confidence of domestic investors but also the stock prices of other offshore listed companies, which resulted in an upsurge in doubts and distrust of the market towards the operation of offshore companies. LaFond and Watts (2008) believe that serious information asymmetry between companies and investors would cause significant decreases in stock prices. When the stock prices of F-1337 were depressed, the TWSE suspended the trading of this company’s stocks and commenced a review of their financial statements. The review results indicated that the disclosure made by the company was complete and true, which finally erased doubts in the securities market in terms of offshore companies. International vulture investors tend to use extreme means (e.g., distributing negative information) to create panic selling pressure in capital markets worldwide. Thus, for the capital market in Taiwan, which is currently in a stage of transformation, encouraging offshore companies for listing can easily turn it to a target of international vulture investors for speculative operations. Berardino (2001) 412
further comments that low criteria for enterprise information disclosure are not only a cause of the Asian Financial Crisis, but also an obstacle for regional economic recovery. Although the TWSE has established references to evaluate the information disclosure levels of listed companies, conservative domestic investors often face challenges when encountering offshore companies with a higher level of information asymmetry. Because domestic investors are comparatively unfamiliar with the operations of offshore listed companies, international vulture investors lure domestic investors by manipulating stocks issued by offshore companies. When vultures attack these offshore companies, the offshore companies are unable to respond or defend themselves in a timely manner because of their location restraint, which results in great losses of the companies and their investors. Therefore, it is essential for investors to utilise available information to make investment decisions when they perceive information asymmetry from offshore companies. Accordingly, we further investigate the investors’ asymmetric responses to the stocks of domestic and offshore companies, which are reflected in stock price fluctuations; moreover, the stock price fluctuations caused by the dramatically rising or falling of VIX are also examined in the extreme stock volatility circumstances. The registration and operation locations of offshore companies listed in Taiwan being comparatively far from domestic investors is a barrier for domestic investors to obtain information. Research by Pirinsky and Wang (2006) shows that domestic investors in the same location as the headquarter of a listed company can gain excess returns when investing in such a company; this verifies the significant correlations between information and geographic location. Coval and Moskowitz (1999 & 2001) find that individual investors prefer to own and trade stocks issued by listed companies with headquarters local to them (this is called “local bias”), because it allows for an information advantage. Ivković and Weisbenner (2005) discover that investors can gain excess returns by trading local stocks if they obtain private information. Research by Massa and Simonov (2006) indicates that local investors have access to local knowledge, which helps them to obtain excess returns by trading stocks issued by local listed companies. Furthermore, other extant studies verify that fund managers and analysts have information advantages when trading local stocks (Loughran and Schultz, 2005; Bae et al., 2008; Ivković and Weisbenner, 2005). Additionally, Chi and Shanthikumar (2017) use IP locations from Google Search to conduct regional analysis and find that local bias can be interpreted as a local information advantage. Listed companies with local bias have a higher bid–ask spread, lower trading volume, and lower responses to earning announcements than those without, which implies that local investors have a local information advantage for investing in listed companies with a higher level of information asymmetry. These empirical results prove that local investors prefer to invest in local listed companies because the information advantage they have may create excess returns. To cope with the rapidly changing investment market, investors consider numerous factors when 413
making investment decisions. News media is the most influential among these factors. As defined in financial theories, mass media and news media are essential channels for investors to access information (Zhang et al., 2017). However, investors may also receive doubtful information. Carroll and McCombs (2003) advise that the number of business bodies run by enterprises affects press coverage and public awareness. In other words, if each business body receives the same quantity of press reports, enterprises running multiple business bodies can create a greater quantity of press reports than those running merely one business body. Danowski et al. (1988) discover that enterprises owning media (e.g., AOL Time Warner and General Electric Company) all receive active attention from their own media, which enables these enterprises to capture investors’ attention and boost their investment sentiment through media manipulation. Bergman and Roychowdhury (2008) further justify that managers would adjust their disclosure behaviour in response to the pessimistic sentiments of analysts. Even when the pessimistic sentiments of analysts have been controlled, the voluntary disclosure of managers still responds to such sentiments, signifying that managers attempt to communicate with general investors as well as analysts. These extant studies indicate that enterprises may utilise certain methods to disclose information for affecting market sentiments. Bulter and Gurun (2012) discover that corporate managers understand the importance of media to investors, and that media may manipulate press coverage on the demand of their parent enterprises. Therefore, enterprises may attempt to use media resources to change investor behaviours, cause fluctuations, and ultimately, boost their stock prices. Zhou (2007) advises that long-term information asymmetry caused by financial news prevents investors from obtaining sufficient information and protection, which easily causes the price of stocks owned by investors to deviate from basic prices. This type of noise-like press coverage could mislead investors into making wrong decisions, resulting in asymmetric responses of positive and negative press coverage towards stock prices. In this study, domestic and offshore listed companies are classified using attributes of companies listed in the TWSE, which helps the researchers to investigate the asymmetric responses of press coverage towards stock prices before and after press coverage. News events and the information that listed companies disclose can generate numerous implications and interpretations, which prevent the use of simple methods to distinguish implications. Therefore, with reference to Manning and Schütze (2002), who define text mining as an effective process of organising unstructured textual information, the researchers of the present study engage in extracting meaningful indices from relevant studies. Subsequently, the researchers of the present study use the text mining technique developed by Sullivan (2001) to explore implicit and meaningful information, processing them into reusable knowledge for storage. Moreover, the support vector machine (SVM) algorithm proposed by Vapnik (1995) can be used for term training to resolve the press coverage categorisation problem. We utilise related press coverage and the excess return (obtained by subtractingmarket return from individual stock return) of the constituents of TSEC Taiwan 50 (TSE50) to conduct term training on both positive and negative news reports, which allows the further categorisation of news about the samples companies into 414
three news variable indicators (positive, negative, and general news) based on the trained news terms. Thus, we establish a financial variable, a corporate governance variable, and a variable that could potentially cause asymmetry in stock prices, thereby facilitating analysis of enterprise responses to stock prices affected by press coverage. Several studies have predicted the default risks of enterprises using the SVM algorithm to classify the financial ratios of enterprises (Huang et al., 2004; Shin et al., 2005; Chen, Hardle, & Jeong, 2010; Chen, Hardle, & Moro, 2011; Trustorff et al., 2011; Shiri et al., 2012). Moreover, Cecchini et al. (2010) use the SVM algorithm to analyse financial statements, which enables them to identify whether managers have committed fraud (Doumpos et al., 2005). Empirical results show that SVMs can precisely identify qualified and unqualified financial statements, which can assist CPAs in auditing. Furthermore, text mining can analyse news material to reveal hidden financial warnings. Research by Lu et al. (2013) suggests that investors can benefit from exploring the forecast signals of business defaults, thereby preventing them from purchasing uncertain investment targets and making unwise investments in doubtful assets. Therefore, this study uses SVMs to classify the attributes of press coverage, which allows us to understand the news trends of domestic and offshore listed companies. Thus, this study analyses the effects of news coverage on the stock prices of domestic and offshore listed companies, thereby exploring the asymmetric responses of press coverage to stock prices. In this study, the materials for news coverage are extracted from the major events of listed markets archived by the Taiwan Economic Journal (TEJ). The research subjects are the foreign enterprises receiving approval from the TWSE for primary listing from January 6, 2009 to March 8, 2016, namely offshore listed companies. We use the SVM algorithm to classify forecasts; thus, the training of positive and negative terms is conducted based on the press coverage and stock prices of the constituents of TSE50. Afterwards, the study uses the trained terms to classify the news coverage of offshore listed companies into three groups: positive news, negative news, and general news. Subsequently, a multiple regression model is used for empirical analysis. Several extant studies utilise the SVM algorithm to classify information, analyse financial information, examine information accuracy, and forecast potential warnings and implications of such information. Therefore, this study also uses SVMs for term training. By adopting and classifying a substantial amount of press coverage, we can then examine domestic and offshore listed companies by news types to investigate stock price changes caused by certain news stories and the effects of information asymmetry on such changes. The empirical results are presented as follows. (1) Compared with the stock prices of domestic companies, those of offshore companies significantly increase on the date that positive news coverage is released. However, the correlation between negative news coverage release and the decline of stock prices is less significant, which signifies asymmetric responses of press coverage to the stock prices of domestic and offshore companies. Therefore, we suggest that the 415
information asymmetry between market investors and offshore companies is greater than that between than market investors and domestic companies. The stock prices of offshore companies exhibit limited decline through increased release of negative news coverage, whereas increased release of positive news coverage generates increases in stock prices. (2) Compared with domestic companies, when the VIX rises significantly, the release of positive news coverage results in slighter increases in the stock prices of offshore companies. When the VIX declines significantly, the release of positive news coverage generates greater increases in the stock prices of offshore companies. Also, when negative news coverage is released, the decline in the stock prices of both domestic and offshore companies is comparatively insignificant. In a time of a great VIX variation, news coverage on domestic and offshore companies generates asymmetric responses in the stock prices of these companies. Because of the high level of information asymmetry in offshore companies, positive news coverage is unable to boost the stock prices when the VIX rises substantially. However, when the VIX declines substantially, the release of positive news coverage brings significant increases in stock prices. The remainder of this paper is organised as follows. Chapter 2 reviews relevant studies and establishes the research hypotheses; Chapter 3 defines the research samples, methods, and related variables; Chapter 4 presents the research results and discussion; and Chapter 5 presents the conclusions and implications. II. Literature Review and Research Hypotheses A. Overseas Companies Listed in the TWSE To enhance domestic economic and industrial development, the Taiwanese government launched several programmes in 2008 (e.g., “1-2-3 Project for Exchange Listing by Offshore Firms,” “Plan on Easing Restrictions on Overseas Enterprises Listing in Taiwan and Appropriately Lifting Restrictions to Allow Mainland Capital to Invest in the Domestic Stock Market,” and “Deregulation on Chinese Qualified Domestic Institutional Investors Investing in Taiwan’s Securities Market,”), which encourage overseas enterprises to list in Taiwan by deregulating certain restrictions in the hope of attracting more overseas enterprises to invest or raise fund in Taiwan. Reese and Weisbach (2001) advise that countries with higher economic development can attract comparatively more foreign enterprises to raise funds and list on the domestic market. Numerous factors exist to urge enterprises to list on the overseas market, which include but are not limited to expanding the overseas market and increasing market shares and enterprise value. Subrahmanyam (1975) suggests that enterprises can break investment barriers by listing abroad. Most overseas listed enterprises tend to apply for listing in nearby countries to raise funds (Pulatkonak and Sofianos, 1999). This situation also applies to foreign enterprises applying for primary listing in Taiwan. Although they register their offices in the British Cayman Islands, the operation and listing of the companies are located in China or nearby Asian countries, thereby allowing them to reduce tax payments, explore overseas markets, and avoid peer competition. 416
Stoughton et al. (2001) propose that enterprises would follow their competitors to list abroad to strengthen their competitive advantage. However, enterprises must carefully assess the investment efficiency of listing abroad, including the comprehensiveness of the laws and regulations of candidate countries and the investment efficiency against taxes. Ashbaugh (1997) argues that enterprises can disclose comparatively more complete information to improve their transparency by listing in countries with solid accounting systems, which can reduce the cost of corporate governance and increase the investment willingness of shareholders. Coffee (2002) declares that enterprises can obtain comparatively low capital costs, reduce the cost of corporate governance, and mitigate agency problems if they choose to list in a market with comprehensive and solid laws and regulations. Amihud and Mendelson (1986) think that enterprises tend to raise funds by listing on a market with high liquidity not only because such liquidity can reduce transaction costs, but also because lower liquidity premiums paid by enterprises reduces capital costs. In addition to comprehensive legal systems and superior market quality, tax preferences are another factor to attract foreign enterprises for listing. Batram and Dufey (2001) indicate that tax preferences can increase the investment willingness of both enterprises and investors, thereby boosting stock liquidity. Moreover, when enterprises are approved for listing abroad, the atmosphere of the market would change accordingly. Cross-listing implies expanding market shares and potential profit guarantees, which can facilitate enterprises in exhibiting their ambition and boosting enterprise value. Therefore, investors understand that excellent decision-making abilities of management reduce agent costs and agency problems, as well as bring optimistic messages of a promising future to enterprises (Pagano et al., 2001). The benefits brought by listing abroad can assist enterprises in growing their business, which also enhances the confidence of foreign listed companies on the stock markets of listing countries. Hargris (1998) advises that a higher level of market openness of a country can lead to positive impacts on the development of its stock market. Bancel and Mittoo (2004) believe that increasing the scale of capital market in a country facilitates enterprises to come into contact with a greater number of investors, thereby attracting overseas enterprises to list. During the current transformation phase of the capital market in Taiwan, the organisational and operational differences between foreign and domestic companies can bring innovative and multiple investment channels to the capital market. Compared with supervisory systems in the United States, Hong Kong, and Singapore, the Taiwanese government imposes different requirements on overseas and domestic companies when they apply for listing. Moreover, information asymmetry is more likely to occur in offshore companies because the government does not make it compulsory for them to include the listing rules in their articles of incorporation. Offshore companies are in essence overseas enterprises whose company registration is located at so-called “offshore financial centres.” Thus, they possess the four characteristics of information asymmetry described earlier. From the perspective of corporate governance, protection of the information of offshore companies is a positive and 417
active information barrier established by tax havens, which increases difficulties in auditing information, whereas tax preferences offered to overseas companies form a negative and passive information barrier that restrains the ability of the government to gather corporate information, thereby resulting in information gaps that affect the stock prices in the investment market. Accordingly, these information barriers affect not only the level of information asymmetry but also investors’ trust in offshore listed companies. Furthermore, the efficiency of financial information disclosure of the MOPS in Taiwan fails to reach investors’ expectations. As the frequency of information asymmetry problems increases, including failure to periodically disclose internal controls, major nonconformity, and improvement updates, more systematic problems related to information asymmetry occur between enterprises and investors. B. Press and Media Since Cutler et al. (1989) examine the correlations between press coverage and stock prices, numerous researchers have explored the influences of press coverage on the investment activities of investors and stock prices. Press coverage is an essential medium that affects the performance of stock prices. Studies have discovered the asymmetric impact of press coverage on stock prices. Press coverage can reduce the level of information asymmetry, and an increase in press coverage is correlated with insider trading activities (Frankel and Li, 2004). If press coverage is assumed to be correlated with stock prices, investors could utilise it to forecast stock price performance. Early acknowledgement of earnings announcements can affect market response, thereby influencing information asymmetry between management and investors (Atiase, 1985). Press coverage is classified by the subjective judgements of people into either positive or negative messages. In rational choice theory, people consider gains and losses to make rational decisions before taking actions. Thus, investors would make trading decisions that benefit themselves, and such decisions reflect stock prices. The trading volume and stock price fluctuations of each investment in stocks shows that interpretations of various investors create different responses. A study by Sankaraguruswamy et al. (2006) shows that private information appear on the announcement date of financial statements, and experienced investors would use such private information to gain profits. Press coverage can affect stock prices and manipulate investment sentiment. This so-called “sentiment” is expectation of market participants in comparison with normal performance; namely, investors who expect stock prices to rise (decline) would anticipate the return ratio to be at a higher (lower)-than-average level. No matter the average level, the existence of such expectations is indisputable (Brown and Cliff, 2004). Consequently, when seller companies own comparatively more private internal information, information asymmetry would occur more easily. Furthermore, the research results by Gentzkow and Shapiro (2010) indicate that the media and press would maximise their own benefits, thereby creating media bias. The meanings and messages implied in press coverage could change enterprise value. Tetlock (2007) and Tetlock et al. (2008) use public news and valuable information related to earning 418
announcements to challenge the efficient market hypothesis. Thus, we believe that the information disclosed in press coverage would create significant fluctuations in stock prices, proving that (1) press coverage would change enterprise value, and (2) investors would utilise press coverage to assess whether their investment activities conform to expectations. Engelberg and Parsons (2011) advise that regional press coverage could forecast trading activities of local investors, which signifies that press coverage can change the decision-making behaviour of investors. Research by Chan (2003) presents the relationships between press coverage and subsequent stock prices, showing that negative news would be reflected in stock prices comparatively slowly. Thus, the present study suggests that investors require more consideration to make investment decisions and responses for negative news because such news brings more risks to them, which postpones their responses to stock prices. Several researchers have investigated this issue in multiple investment markets. Griffin et al. (2011) examine the influences of news release on the stock prices of 56 securities markets worldwide. The research results indicate that stock prices in developed markets would be affected on the date of news release, but stock prices in emerging markets would be affected on both the date of news release and of non-news release. However, the effectiveness of press coverage can be influenced by enterprises because certain enterprises may have control over the press and media, thereby urging them to release slanted press coverage. After controlling all known risk factors, Fang and Peress (2009) found that the returns of stocks without press coverage are higher than those with high press coverage (i.e., receiving monthly press coverage of 0.65–1%). This is even more obvious on the stocks of small companies, with high ownership, with few analysts following-up, and with high stock volatility. Accordingly, press coverage may not only imply information that affects stock prices, but also the existence of unknown hidden messages. Zhou (2007) suggests that asymmetric financial press coverage prevents investors from gaining sufficient information and protection. Although the purpose of press coverage is to disclose information, different press coverage generates new problems that in turn, affect investment activities on the stock market. Barber and Loeffler (1993) investigate the influences of second-hand information provided by analysts in the investment column of the Wall Street Journal on stock prices and trading volume; the research results indicate that the stocks recommended by the column brought 4% of excess returns and the average trading volume of such stocks increased two-fold in 2 consecutive days. However, regarding the response asymmetry of stock prices caused by varying press coverage, Tetlock (2007) utilises the daily news of the Wall Street Journal to verify that the negative substantial information represented by pessimistic press coverage causes stock prices to decline. Das (2007) develops an algorithm to extract the sentiments of individual investors from stock message boards; the empirical results indicate that investor sentiment is significantly correlated with stock prices. Therefore, this study suggests that such a correlation is caused by journalists who write 419
the news reports; specifically, if journalists have positive (negative) thoughts, the news reports they write would carry a positive (negative) perspective. Dougal et al. (2012) explore the correlations between financial news reports in the Wall Street Journal and the performance of stock prices; the results verify that financial journalists are capable of reinforcing and mitigating investor sentiment. Therefore, press coverage can directly and indirectly affect stock prices and investor behaviours. C. Research Hypotheses Based on the research scale of the literature review, the operations and competitiveness of offshore companies are not inferior to those of domestic companies. However, the levels of information disclosure and transparency in offshore companies are weaker than those of domestic companies, which results in information asymmetry. Moreover, the messages revealed by press coverage affect investor sentiment and stock prices. Tetlock (2007) asserts that press coverage would directly affect investors’ opinions on stocks. The research results of Barber and Odean (2006) verify that individual investors are more easily affected by press and media than institutional investors. After comparing the behaviour of different types of investor, their study further identifies that general investors tend to buy the stocks that have greater press coverage. In other words, when investors face multiple investment choices, they tend to buy stocks that attract more attention. This situation applies to the stock market in Taiwan, where the majority of investors are individual ones. Thus, compared with actual operational performance, the press coverage of listed companies could have a greater influence on stock prices. Accordingly, the present study proposes four hypotheses to examine (1) the asymmetric responses of positive and negative news coverage on the stock prices of offshore and domestic companies, and (2) the influences of the VIX on the stock prices of offshore and domestics companies when positive and negative news is released. Mitton (2002) finds that high corporate information transparency can enhance the performance of stock prices, whereas a low level of information transparency impairs the performance of stock prices. Therefore, the present study proposes the following two hypotheses to examine the influences of press coverage on the stock prices of offshore and domestic companies under the condition of information asymmetry. H1: When positive news is released, the stock prices of offshore companies rise more significantly than those of domestic companies. H2: When negative news is released, the stock prices of offshore companies decline more significantly than those of domestic companies. The Chicago Board Options Exchange (CBOE) is the world’s largest options exchange; thus, the types of trading product and the trading volume of options can serve as indicators. The VIX proposed by the CBOE is the most representative indictor for interpreting investor sentiment. 420
Because the VIX reflects the cost that investors are willing to pay to handle investment risks, such an index can also reflect the panic level of investors regarding the outlook for the stock market. Whaley (2000) finds that the VIX has a negative correlation with the stock index of the market. A rise in the VIX means an increase in the investment risks of stocks, which increases the level of investor sentiment. Therefore, the researchers of the present study propose two hypotheses to examine the volatility in the stock prices of offshore and domestic companies when investor sentiment is affected by information asymmetry and the release of news. H3-1: When the VIX increases significantly, the release of positive news causes the stock prices of offshore companies to climb less significantly than those of domestic companies. H3-2: When the VIX declines significantly, the release of positive news causes the stock prices of offshore companies to climb more significantly than those of domestic companies. H4-1: When the VIX increases significantly, the release of negative news causes the stock prices of offshore companies to decline more significantly than those of domestic companies. H4-2: When the VIX declines significantly, the release of negative news causes the stock prices of offshore companies to decline less significantly than those of domestic companies. III. Research Methods News has multiple meanings and can be interpreted from various perspectives. This study uses the SVMs of text mining as the algorithm for term training, which extracts meaningful indicators from relevant texts. We further utilise the text mining technique proposed by Sullivan (2001) to define it as a process of editing, organising, and analysing a large amount of documents to extract critical messages, thereby exploring hidden textual characteristics in the news. Term training of positive and negative news is conducted using the excess returns (i.e., the return of individual shares minuses the return of the market) of TSE50 constituent stocks and the corresponding press coverage. The trained news terms are used to classify the press coverage of the sample companies into three groups: positive news, negative news, and general news. The three groups serve as three news variables to examine the correlations between positive or negative press coverage and the stock returns of offshore and domestic companies. A. News Classification 1. Methods of News Term Training Data preprocessing is conducted using the Chinese term segmentation system developed by the Chinese Knowledge and Information Process Team of Academia Sinica to analyse the trained news terms and the news corpus of the sample companies. After the news corpus is categorised into the sets of Chinese terms, noise terms that cannot highlight the meaning of the news corpus are removed, such as “(comma),” “of,” and “due to.” Therefore, the trained news corpus turns into the sets of terms to replace the original type of continual texts. This paper refers to the research methods of Cecchini et al. (2010) to classify the news corpus into three groups: positive 421
news, negative news, and general news. Positive news is defined as news that makes the stock price rise at least 1% higher than the market returns on the date of its release. Negative news is defined as news that makes the stock price drop at least 1% lower than the market returns on the date of its release. The rest of the sampled news is classified as general press coverage. Based on the categorisation results of the news corpus, three categorisation codes of positive terms, negative terms, and general terms are added in front of the terms to complete the term training of categorisation labelling. Information types of the same trained terms may have three labels (positive, negative, and general terms) or any two of these labels, because the segmentation is conducted on different companies and news items. Therefore, this study uses an objective statistical categorisation method to assist in categorising term types. Cortes et al. (1995) use statistical learning theory to develop an innovative machine learning method—SVMs—which is an appropriate model for the categorisation and regression of data. SVMs can create a model through training, and such a model can predict the type of unclassified data. In other words, SVMs can locate a decision boundary on the high dimensional space of a hyperplane to divide two types of data sets. 2. Categorising Positive and Negative news After processing the segmentation system, this study selects certain terms as features of the news corpus using a statistical method to calculate the proportion that each term represents within the news corpus. Common strategies of feature proportion selection include a binary pattern (i.e., determining whether certain terms have appeared), or the term frequency (TF) or document frequency of certain terms. The proportion of terms is calculated, and the terms with a lower proportion are then eliminated; thus, the rest of the term sets become the features of the corpus. We utilise TF to acquire features. Frakes and Baeza-Yates (1992) also use TF to extract features, considering that terms with high-frequency terms have higher levels of relevance with the subject categories of a corpus. In other words, a high-frequency term increases has a high level of relevance with the features of specific subject categories. The value of TF can be presented as follows: (������������������������: the appearance times of term j in the subject category i (i = 1 [positive features], 2 [negative features], and 3 [general features]). ������������������ = ∑������������=1 ������������������������ ; ������ = 1– ������, feature i has m types of terms The tf value of all terms in each piece of the corpus is calculated and added according to their subject categories. The subject category that obtains the maximum total of tf value is then deemed as the representative categorisation of relevant pieces. In doing so, this paper effectively categorises all news items of sample companies as positive news, negative news, or general 422
news. B. Regression Model to Forecast Company Stock Returns Using Press Coverage To examine the hypothesis on the influences of positive and negative news on stock prices, a multiple regression model is used to review the significant changes in stock returns after the release of the positive and negative news. The dependent variable is stock returns (RET), and the independent variables are positive news, negative news, and control variables. The regression model is presented as follows, where i represents the sample companies, and t represents the cumulative days of stock returns. ������������������������,������ = ������0+������1������������������������������������������������ ������������������������������,������ + ������2������������������������������������������������ ������������������������������,������ + ������������������������������������������������������ ������������������������������������������������������������,������ + ������������,������ (1) C. Regression Analysis on the Cumulative Returns of Stock Prices To examine the hypothesis on the influences of positive and negative press coverage on offshore and domestic companies, a multiple regression model is used to verify significant changes in stock returns of offshore and domestic companies after the release of positive and negative news. The dependent variable is the excess return on stock prices (RET), and the independent variables are positive news and negative news. Positive news and negative news are dummy variables that take values of either 1 or 0. When positive news is multiplied by offshore companies, and the company serves as a dummy variable; the offshore companies’ value is 1; otherwise, it is 0 and the control variable. The regression model is presented as follows, where i represents the sample companies, and t represents the cumulative days of stock returns. ������������������������,������ = ������0+������1������������������������������������������������ ������������������������������,������ + ������2������������������������������������������������ ������������������������ ������ ������������������������������������������,������ + ������3������������������������������������������������ ������������������������������.������ + ������4������������������������������������������������������ ������������������������ ������ ������������������������������������������,������ + ������������������������������������������������������ ������������������������������������������������������������,������ + ������������,������ (2) D. Regression Analysis on Stock Returns Affected by News Variables and VIX To examine the hypothesis on the influences of VIX on the fluctuations of stock prices, a multiple regression model is used to verify significant changes in the stock returns affected by the release of positive and negative press coverage and VIX variation. The dependent variable is the excess returns on stock prices (RET), and the independent variables are positive news and negative news. Positive news and negative news are dummy variables, and their values are either 1 or 0. When positive news and negative news are multiplied by VIX variation, VIX variation serves as a dummy variable; specifically, the value of VIX is 1 when it rises and 0 when it drops. The control variables are the standard unexpected earnings (SUE), monthly cumulative returns in the preceding period, VIX, VIX variation, trading volume, market value, price/book ratio (PBR), and beta value. The regression model is presented as follows, where i represents the sample companies, and t represents the cumulative days of stock returns. RET������,������ = 423
������0+������1������������������������������������������������ ������������������������������,������ + ������2������������������������������������������������ ������������������������ ������ ������������������������������������������,������ + ������3������������������������������������������������ ������������������������ ������ ������������������������,������ + ������4������������������������������������������������������ ������������������������������,������ + ������5������������������������������������������������������ ������������������������ ������ ������������������������������������������,������ + ������6������������������������������������������������ ������������������������ ������ ������������������������,������ + ������������������������������������������������������ ������������������������������������������������������������,������ + ������������,������ (3) IV. Research Samples A. Training Sample Classification This study utilises the press coverage of the constituent stocks of TSE50 and the corresponding stock returns to establish the training data for machine learning. This is because the constituent stocks possess certain representativeness in Taiwan’s securities market. To serve as subject stocks for training, the news corpus is extracted from the major events of 50 constituents archived by the TEJ. The news collection time frame is from 2009 to 2016, and the total number of times the constituents appear in press coverage is approximately 110,000. We use the segmentation system proposed by Academia Sinica to eliminate noise terms, thereby acquiring approximately 2.6 million terms. This study refers to the research method proposed by Cecchini et al. (2010) to categorise the news corpus into three groups: positive news (17%), negative news (20%), and general news (63%). We add three categorisation codes (positive, negative, and general) in front of the terms, thereby dividing all samples into two parts: the training set and backtesting set. To ensure the final categorisation of all terms, the categorisation model using SVMs is utilised to establish the training phase and backtesting phases. B. Research Samples of Offshore and Domestic Companies To investigate the press coverage of the offshore and domestic companies listed on the TWSE, the researchers collect major event news items archived in the TEJ from January 6th, 2009 to March 14th, 2016 to serve as the research variables. The research samples of 874 listed companies include 53 offshore companies and 821 domestic companies. C. Research Variables 1. News Variable To investigate whether the release of news on offshore companies would affect investor confidence,we must define the variables precisely to capture the influence levels on stock returns. Therefore, this study uses the two research variables of negative news and positive news, which are described as follows. (1) Positive news Variable This research utilises positive news as a dummy to assign individual news stories with a value of either 1 (positive news) or 0 (otherwise). We can refer to this to determine the outlook of the stock prices on the market. A high proportion of positive news for a company means that it has 424
more positive information, resulting in improved perceptions of investors. {���������O���������t���h���������e���������r������w��� ���i���s���e���������, ������ ,1 } 0 (2) Negative News Variable This research utilises negative news as a dummy to assign individual news stories with a value of either 1 (negative news) or 0 (otherwise). We can refer to this to determine the outlook of the stock prices on the market. A high proportion of negative news for a company means that it has more negative information, resulting in worsened perceptions of investors. {���������������O������t���h������e���������r������w��� ���i���s���e������,���0������, 1 } 2. Control Variables Subsequently, we refer to the information that would affect the stock returns of offshore and domestic companies to select the financial determinants as control variables, which are the SUE, monthly cumulative returns in the preceding period, VIX, VIX variation, trading volume, market value, PBR, and beta value. Because trading volume and market value are much larger than the other determinants, natural logarithms are extracted for calculation convenience. The control variables are defined as follows. (1) SUE The SUE equation proposed by Chan et al. (1996) and Chordia and Shivakumar (2005; 2006) shows the process of standardisation, in which the unexpected earnings is divided by the standard deviation of the last eight quarters: ������������������������,������ = (������������������������,������ − ������������������������,������−8)/������������������ ������������������������,������ : The latest announced quarterly earnings per share (EPS) of Stock j after Month i ������������������������,������−8 : Mean of Quarterly EPS of Stock j during the eight quarters preceding Month i ������������,������ : Standard Deviation of Quarterly EPS of Stock j during the eight quarters preceding Month i (2) Monthly Cumulative Returns in the Preceding Period (Cumulative Returns) We use the stock returns for 90 cumulative trading days until 2 days before press coverage (i.e., the investment period) as the substitution variable to evaluate the stock performance of the listed companies. ������������,������,������: total stock returns of Stock i during the investment period (t, T) (3) Trading Volume Karpoff (1987) claims the trading volume is positively correlated with the absolute value of stock price fluctuations. Therefore, the present study uses trading volume as a control variable. Because the value of trading volume (unit: million shares) is too large, the trading volume in this study is processed using the natural logarithm. 425
(4) Market Value Because the actual value of companies cannot be accurately calculated, market value is used to reflect current enterprise value as well as the value of corporate stocks on the securities market. Market value is negatively correlated with financial crises and stock prices (Brazel et al., 2009). Because market value (unit: million NTD) is a large number, the market value in this study is processed by using the natural logarithm. (5) PBR Garlappi and Yan (2011) suggest that a PBR can be used to capture financial crises. Companies that are regarded as having inferior outlook or low growth on the market generally have a higher PBR or lower stock prices. Compared with companies with high stock prices and a low PBR, such companies have a higher probability of experiencing financial crises and possess higher capital costs. Therefore, the present study assumes that PBR would have a positive correlation with the probability of financial crises. (6) Beta Value (Beta Coefficient) Beta value reflects investor sentiment towards the stock market. When it increases, the volatility of stock returns rises during the upturn of the market, whereas when it decreases, the volatility of stock returns drops during the downturn of the market. Kumar and Persaud (2001) indicate that investors would reduce risk assets when their risk demands decrease, thereby jointly reducing the value of risk assets. When the risk demands of investors increase, an increase in the demand of risk investment would cause the value to rise. Because the expected returns on stocks would be affected by systematic and unsystematic risks, the actual returns would become unstable. The returns on individual assets would be affected by system risks, and hence, a beta value is used to represent the level of influence. In other words, a beta value is used to present the fluctuation of the expected returns on individual assets when market returns fluctuate. V. Empirical Results A. Means Test and Descriptive Statistical Analysis The descriptive analytical results of the research samples are presented in Table 1, including the research year, number of news samples, pre- and post-changes in stock returns, news variables, and control variables (e.g., mean, SD, minimum, maximum, skewness, and kurtosis). Panel A in Table 1 presents the descriptive analysis of the news variables and the changes in stock returns before and after news release. The mean is 0.24 on the day before the news of stock returns is released, and drops to 0.06 on the release date. However, regarding cumulative stock returns, the mean is 0.12 on the day after the news is released, and increases to 0.29 on the 7th consecutive day. Panel B in Table 1 presents a descriptive analysis of the control variables. The means of negative and positive news are 0.21 and 0.22, respectively; the mean of SUE is 0.026; the mean 426
of the cumulative returns is 3.53; the means of the VIX and VIX variation are 22.04 and -0.00011070, respectively; the mean of the market value is 105,081; the mean of the PBR is 1.91; and the mean of the beta value is 0.918. B. Means Test of Offshore and Domestic Companies The results of the means test exploring the influences of positive and negative news on the stock returns of domestic and offshore listed companies are presented in Table 2. A significant positive correlation is observed between stock returns and positive news from the day before the release of positive news (2.68) to the sixth day after its release (2.18). However, a significant negative correlation is determined between stock returns and negative news from the fourth day after the release of such news (-2.88) to the sixth day (-2.06), which indicates that compared with positive news, the impact of negative news on stock returns is less significant. Thus, H2 is not supported. Moreover, an analysis of differences reveals that the positive correlation between stock returns and positive news is observed from the day before the release of news events to the sixth day after its release, which highlights that the increase in stock prices generated by positive news of offshore companies is higher than that of domestic companies. Thus, H1 is supported. C. Regression Analysis on the Stock Returns of Offshore and Domestic Companies Affected by News Variables Results of the regression analysis performed to determine the influences of press coverage on the stock returns of domestic and offshore companies on the day before and after news release are presented in Table 3. Negative news about offshore companies has less correlation (-1.39), which means that investors have perceived high information asymmetry in offshore companies listed on the market, and are thus less affected by negative news. Because the negative impacts of negative news are insignificant, H2 is not supported. Positive news about offshore companies exhibits a significant positive correlation with stock price (3.52), which indicates that the positive effect of positive news exceed investors’ expectations. Because positive news results in superior returns, H1 is supported. Furthermore, negative news about domestic companies shows a significant negative correlation (-26.74), and positive news about offshore companies shows a significant positive correlation (15.63). Additionally, from the total results of the analysis, negative news has a significant negative correlation with domestic companies (-26.66), whereas negative news has no significant correlation with offshore companies (1.19), indicating that negative news has limited influence on offshore companies. Moreover, positive news has a significant positive correlation with domestic companies (15.65), whereas positive news has a positive correlation with offshore companies (2.74), entailing that investors have perceived high information asymmetry in offshore companies, and the difference in investors’ perceptions of these two types of companies are reflected in stock returns. Accordingly, the press coverage of domestic and offshore companies could cause asymmetric responses to stock returns. 427
IV. Regression Analysis on Stock Returns by Multiplying News Variables and VIX Variation VIX reflects how much investors are willing to invest to confront investment risks. Thus, VIX variation reflects the fluctuations in investor sentiment towards the outlook of the market. A high VIX means that investors feel anxious about the stock market, whereas a low VIX means the fluctuations in the stock market are comparatively mild. The total analysis in Table 4 indicates that the negative news of domestic companies has a significant negative correlation with the VIX variation (-6.12), which means that negative news generates more negative impacts than investor expectations. The result shows a high level of investor anxiety, which leads to comparatively fierce stock price fluctuations on the market after the release of negative news. Negative news about offshore companies has no correlation with the VIX variation (-0.73), which means that the acceptance of investors offsets their anxiety when they perceive panic, thereby minimising the impacts of negative news and resulting in no significant difference. Thus, H4-1 is not supported by the results. Moreover, positive news about domestic companies has no correlation with the VIX (-1.17), and positive news about offshore companies has a significant negative correlation with the VIX (-2.56), indicating that the market is under anxiety when the VIX rises, thereby restricting the effect of positive news on stimulating stock prices. Thus, H3-1 is supported by the results. V. Regression Analysis on Cumulative Stock Returns by Multiplying News Variables and Domestic and Offshore Companies Table 5 enhances understanding of changes in cumulative stock returns by showing that stock returns have significant negative correlations with the negative news of domestic companies from the day before news release (-19.58) to the sixth day after the release (-17.13), indicating that the influences of negative news on stock returns may last over a week. Moreover, the cumulative stock returns of offshore companies show nonsignificance from the day before to the fifth day after the release of negative news; it began to show a significant positive correlation only since the sixth day (2.08). This suggests that a clear difference would only be observed in the response to negative news after approximately 1 week of the news release. Thus, H2 is not supported by the results. Positive news about domestic companies shows significant positive correlations, whereas the positive news of offshore companies shows significant positive correlations from the day before news release (3.96) to the second day after (2.06), changing to insignificance afterwards. This indicates that positive news generates positive responses towards stock returns on the first 3 days, and then shows adjustment to the previous levels. VI. Regression Analysis on Cumulative Stock Returns by Multiplying News Variables and VIX Variation As shown in Table 7, the researchers add VIX variation to determine the impacts on stock returns when investors face anxiety in the market, which enables exploring the correlations between 428
press coverage and VIX variation. Negative news about domestic companies shows significant negative correlations from the day before news release (-7.54) to the sixth day after (-4.19), which indicates that the negative impacts of negative news would continue for more than a week when the market is under anxiety. However, despite the significant correlation on the fifth day after news release (2.16), negative news about offshore companies shows no correlation with VIX, indicating that negative news would not generate excessive negative impacts to increase market anxiety. Thus, H4-2 is not supported, which signifies the asymmetric responses of press coverage to the stock prices of domestic and offshore companies. Moreover, positive news about domestic companies has negative correlations with VIX variation from the day before news release (-2.36) to the second day after (-2.91) with no correlation afterwards. Positive news about offshore companies has negative correlations with VIX variation between the day before news release (-2.08) and the second day after (-2.13) with no correlation afterwards. Thus, H3-2 is supported by the results. This highlights that domestic and offshore companies have asymmetric stock prices from the day before news release to the second day after when they face positive news and VIX variation. However, the stock returns show adjustment to previous levels from the third day after news release. The research results of this study verify the information asymmetry in offshore companies. While investors are affected by news information, they still have concerns about the offshore companies. Therefore, compared with the stock returns of domestic companies, negative news would lead to less significant influences on the stock returns of offshore companies. Regarding the influences of news information and VIX variation on the stock returns of offshore companies, the statistical results clearly indicate that the release of positive news would trigger greater influences than negative news would. Therefore, investors are suggested to carefully evaluate the asymmetry between news information and stock returns to decide on investing in offshore companies. VI. Conclusions To examine the asymmetric responses of press coverage on the stock prices of domestic and offshore companies listed in Taiwan, we use news variables to analyse correlations between information and stock price fluctuations, thereby exploring the impacts of information asymmetry on investor sentiment and stock prices. The empirical results of this study are as follows. (1) Regarding the stock prices of offshore companies on the day of news release, positive news could urge stock prices to rise significantly on that day; however, the decline caused by the release of negative news is insignificant. Thus, press coverage causes asymmetric responses to the stock prices of domestic and offshore companies. (2) When the VIX substantially rises (drops), the release of positive news causes the stock prices of offshore companies to rise less (more) significantly than those of domestic companies. However, the declines caused by negative news are insignificant. Thus, when fluctuations in the VIX are 429
conspicuous, press coverage would cause asymmetric responses to the stock prices of domestic and offshore companies. Because of a higher level of information asymmetry in offshore companies, positive news is unable to stimulate rises in stock prices when the VIX increases drastically. However, when the VIX drops drastically, positive news can cause stock prices to surge significantly. The empirical results indicate that offshore companies can gain more stock returns from positive news than domestic companies can. Because of the higher information asymmetry of offshore companies, investors have relatively low assessment on their stocks, which is reflected on the stock prices. However, positive information generated by offshore companies would encourage investors to expect profits to grow markedly in the near future. From the perspective of investors, positive news is a guarantee of future profits. Because the release of positive news verifies the aforementioned guarantee, increased investor confidence would reflect on stock prices directly, thereby boosting the growth of offshore companies’ stock prices more drastically than those of domestic companies. The empirical results comply with the effective market hypotheses proposed by Tetlock (2007) and Tetlock et al. (2008), which verifies that press coverage can change the enterprise value of listed companies. On the contrary, because investors have already lowered the assessment on offshore companies because of information asymmetry, negative news does not result in extraordinary responses. LaFond and Watts (2008) identify similar results in that asymmetry between listed companies and traders would make stock prices decline; thus, based on this known premise, investors would reasonably perceive such asymmetry, thereby reducing the responses of stock prices. Studies have explored the correlations between press coverage and stock prices (Chan, 2003; Tetlock et al., 2008; Fang and Peress, 2009; Griffin et al., 2011; Dougal et al., 2012). Press coverage of the securities market has affected investor behaviour, thereby driving stock prices to hike or slump. This indicates that press coverage can forecast stock prices, thereby influencing the performance of stock prices indirectly. Based on such a perspective, we classify the attributes of companies listed on the securities market to further analyse the influences of positive and negative news on stock price fluctuations. The empirical results show that negative news would affect stock returns for a comparatively long period than positive news would. Moreover, domestic companies would be more significantly influenced by press coverage and VIX variation than offshore companies. These empirical results can serve as a reference for investors to make investment decisions. To become the financial centre of the Asia-Pacific region, Taiwan’s capital market has actively made efforts to attract foreign institutions to invest to accelerate the market’s development and the mature operation of the real economy in Taiwan, including lifting the ban on foreign enterprises to list and issue TDRs in Taiwan. However, several events (e.g., international vulture 430
companies attacking listed offshore companies and sudden bankruptcy announcements of TDR companies) have urged the market to examine the comprehensiveness of its management mechanisms and regulations to prevent such events from occurring again. Although the government would implement administrative punishments such as trading suspension, delisting, and penalties against listed companies that violate legal regulations and laws, no criminal punishment is imposed, which increases distrust from investors. This indicates that room exists for improving the domestic securities market in Taiwan. Therefore, the government shall urge listed companies to improve information transparency by disclosing true and complete information. In addition, to more effectively realise the trading principle of fairness on the capital market, the government shall impose criminal punishments against companies that disclose false and incomplete information. 431
ISSSM-0278 The Impact of Reverse Merger on Stock Price Behavior – Evidence in Taiwan Market Shao-Huai Liang, Yu-Ting Hsieh, Hsuan-Chu Lin and Hsuan-Yi Ho National Cheng Kung University E-mail: [email protected] Abstract Reverse merger makes it possible for private companies to go public through merging of listed shell firms. This method is completely different as compared to the traditional IPO process. Private firms can achieve the purpose of going public at a relatively quick rate that is relatively low cost using reverse merger. In this research, 58 reverse merger cases in the Taiwan market from 2007 to 2017 were collected. Of the 58 shell companies, 21companies are TWSE-listed, and 37 companies are TPEx-listed. In addition, 40 shell companies changed to different industries after being merged by private firms, whereas the remaining 18 shells remained in their original industries. Our research result shows that shell companies may be the target in reverse mergers mainly because of their poor financial performance and operating conditions before being merged. As a result, the original management may has no intention to continue operating the companies. Additionally, we also find that the stock price of shell companies that change their industries after reverse mergers will cause a fluctuation phenomenon compare to other similar public firms. Furthermore, we utilize the event study method to examine the abnormal returns surrounding reverse mergers. We find that a positive abnormal return exists both before and after the announcement date of a reverse merger. Keywords: Reverse Merger, Financial Performance, Stock Performance, Abnormal Return 432
ISSSM-0374 Firms Green Energy Investment Strategy with Carbon Tax and Subsidy Policy - Analysis of Real Options Approach Chuan-Chuan Koa, Chien-Yu Liub, Jing Zhouc, Zan-Yu Chend, Hong-Tao Jiange a,c,d,e Department of Business School, City College of Dongguan University of Technology, Guangdong, China b Department of Food and Beverage Management, Jinwen University of Science & Technology, New Taipei City, Taiwan E-mail: [email protected] a, [email protected] b, [email protected] c, [email protected] d, [email protected] 1. Background This paper proposes that the United Nations Intergovernmental Panel on Climate Change (IPCC) must maintain global warming within the controllable range. Its core job is to price carbon emissions. In order to achieve CO2 net emissions, the goal of “net zero” must be achieved. This paper mainly constructs the green energy investment strategy model of firms in response to government environmental protection policies. From the perspective of social welfare, when the government levies an CO2 emission tax on firms, in order to prevent firms from transferring the increased costs to consumers and reduce social welfare, it is necessary to provide subsidy to reward carbon-reducing firms. Encourage firms to invest in green energy equipment and innovative production technology. This paper uses the Real Options Approach (ROA) to evaluate the investment decision model. Mostly, the Net Present Value Methods (NPV) is used for the feasibility evaluation of investment projects. However, in a complex and uncertain investment environment, the manager’s investment strategy should adopt a dynamic decision analysis model. Among them, the ROA is more responsive to the complex investment environment than the traditional NPV (Myers,1977; Dixit and Pindyck, 1994). Torani et al. (2016) used the stochastic dynamic model of the ROA to explore the changes in solar technology. In the case of uncertain electricity prices and solar energy costs, assess energy, electricity prices, innovation subsidies and carbon taxes, as well as the optimal investment threshold for enterprises, and propose policy development as the main key factor. Ansaripoor and Oliveira (2018) constructed a model using the ROA of management flexibility to calculate uncertain CO2 prices, fuel prices, fuel consumption, and technological advances. Then, analyze how uncertain fuel prices and technological advances affect the value of option. 2. Methods This paper uses the ROA of dynamic investment decision-making to construct a green energy investment strategy model for firms to respond to the government's carbon emission policy. 433
Under uncertain carbon emissions, assuming that the quantity of tax Pe t and subsidy tax Pe t per unit of CO2 emissions are uncertain variables, the change follows the joint geometric Brownian motion. H indicates the ratio of the government's actual levies CO2 emission tax and reduction of CO2 emission subsidies. Where H* Pe t* / Pc t* , upgraded the green energy equipment for the firms, the government’s threshold for CO2 emission tax and reduction of CO2 emission subsidies ratio. Then the value function of the firms to update the production equipment is: F (H ) -A1H 1 , H H * Q1e Q2e (1) Q2e H R I g- r e r c , H H* r To solve H* 1 r e R r c Q1e Q2e r Ig r r c and A1 Q2e H*11 based Q2e r r c 1 1 1 r e on the value matching condition and smoothing condition (Dixit and Pindyck, 1994) from Eq.(1). 3. Results This paper mainly uses the ROA to construct a policy strategy model from the perspective of social welfare. The optimal of the threshold H * for CO2 emission tax and reduction of CO2 emission subsidies ratio. If H is higher than the threshold H * , the firms will suspend investment in green energy equipment because of insufficient investment subsidies. If the H is less than or equal to the threshold H * , the firms will receive emission subsidies and the firms should invest in green energy equipment. The results of this study can provide reference for firms to invest in green energy equipment and government control of emissions policies. To achieve co-construction, co-governance, and sharing innovative social governance, and to create a win-win situation for the government, firms and society. Keywords: Real options, Carbon tax, Subsidy policy, Uncertainty, Social Welfare Acknowledgments The authors would like to thank the City College of Dongguan University of Technology of the China, for financially supporting this research under Contract No. 2017YZDYB03R. 4. References Ansaripoor A. H., & Oliveira F. S., (2018). Flexible lease contracts in the fleet replacement problem with alternative fuel vehicles: a real-options approach. European Journal of Operational Research, 266, 316-327. Dixit, A.K. & Pindyck, R.S., (1994). Investment under Uncertainty, New Jersey: Princeton University Press. Myers, S.C., (1977). Determinants of corporate borrowing. Journal of Financial Economics, 5, 434
146-175. Torani, K., Rausser, G., & Zilberman, D., (2016). Innovation subsidies versus consumer subsidies: a real options analysis of solar energy. Energy Policy, 92. 255-269. 435
ISSSM-0370 Factors Affecting Customer Satisfaction of Using Mobile Banking Services in Ho Chi Minh City, Vietnam Keng-Shen Chena,*, Ming-Kun Lina, Mei-yu Hsub a Dept. of Finance, Lunghwa University of Science and Technology, Taiwan R.O.C. b Dept. of Marketing Distribution Management, Chien Hwsin University of Science and Technology, Taiwan R.O.C. * E-mail: [email protected] Abstract Mobile phones have created a platform to expand commercial transactions in a very easy manner and have created a wide array of business opportunities through the expansion of wireless communication. These developments facilitate business transactions, trading, and purchasing of goods and services without much effort. Therefore, this study is based on the previous literatures to investigate factors affecting customer satisfaction of using mobile banking services in Ho Chi Minh City, Vietnam. In this study, Technology Acceptance Model (TAM model) is used to explain the customer satisfaction to use mobile banking services. Sample data were collected from people who have experienced of using mobile banking services. The empirical results from demographic analysis, reliability analysis, factor analysis, Pearson correlation test and regression analysis showed that perceived usefulness, perceived ease of use, relative advantage and perception of risk impact on customer satisfaction of using mobile banking. In addition, from ANOVA analysis, it found that gender is a significant factor to use mobile banking. The contribution of this study can develop awareness of customer about perceived usefulness of mobile phone. Customer can give the feedback to manager easily and quickly via mobile phones. Keywords: mobile banking, technology acceptance model, customer satisfaction, Ho Chi Minh City, Vietnam 1. Background Mobile banking (MB), which is also known as cellphone banking is “the use of mobile terminals such as cellphones and personal digital assistants (PDAs) to access banking networks via the wireless application protocol (WAP)” (Zhou et al., 2010). As one kind of Internet banking, MB arranges for faster and more convenient method than original banking. To take part in MB, users only need a device such as a mobile phone with internet connection. Within some simple account registering provided by the banking institution, the participants would easily transact money from anywhere. There are several procedures to access MB, such as by web browsing, messaging, or application (Board of Governors of Federal Reserve Systems, 2012). Three main fundamental transactions of MB are storing money, completing cash-in and cash-out transactions, 436
and transferring money. The aim of this research is therefore to find out the key factors influence on customer satisfaction on using mobile banking services in Ho Chi Minh City, Vietnam. Mobile Banking (MB) is among the latest in a series of recent mobile technological wonders (Mehrad & Mohamm, 2015). Although Automated Teller Machine (ATM), telephone, and internet banking offer effective delivery channels for traditional banking products, but as the newest delivery channel established by retail and microfinance banks in many developed and developing countries, MB is likely to have significant effects on the market (Mehrad & Mohamm, 2015). Since the number of cell-phone is more than personal computer (PC), MB has become more popular than e-banking among bankers. Also, mobile phones enhance the quality of services because clients can perform their financial jobs in every time and place. Therefore, it is clear that use of mobile phones for banking problems is useful for both clients and the bank. This leads to the establishment of a stronger relationship between the financial institutions and clients (Laukkanen, 2007). Despite such benefits, the use of mobile phones or tablets to conduct banking transactions or access financial information is not as widespread as might be expected (Dineshwar and Steven, 2013). MB appeared in Vietnam in 2003, but so far, most commercial banks only use the SMS channels to query general information and bank account information. Although the function of payment/transfer on the MB channel was developed in 2006 but until December 2014 there are 32 banks provide these services (baomoi.com). In general, MB is not a common payment channel in Vietnam. MB did not show very exciting signals either. As e-banking moves slowly and cautiously forward, MB in Vietnam has only made a few baby steps. The current usage of MB in Vietnam is below its potentials. According to a representative of Smartlink card services joint stock Company, the growth of MB in Vietnam need more time to meet its demand because customers are still used to their familiar payment channels. Another difficulty many banks meet when they started providing MB services is the payment habit of Vietnamese. This study is to study factors affecting customer satisfaction of using MB services in Ho Chi Minh City. The main objective of this work is to achieve a more well understanding of factors influencing the customer satisfaction of using mobile baking services. There are seven divided objectives would be done in this research as following: Firstly, this study measures shows perceived usefulness could influence customer satisfaction on MB services. Secondly, this study investigates the effect of perceived ease of use to customer satisfaction on MB services. Thirdly, this study also figures out how relative advantage to affect to customer satisfaction on MB services. 2. Methods 2.1 Research Model 437
Perceived usefulness is the degree to which a person believes that using a system would enhance his or her job performance (Mehrad and Mohammadi, 2015). Perceived usefulness has been identified as having a significant positive correlation with both attitude and usage intention, for instance, perceived usefulness positively influences the adoption of mobile internet and mobile services (Shaikh and Karjaluoto, 2015). Many prior studies have shown that perceived usefulness have a direct dramatic impact on behavioral intention to use an online system (Hanafizadeh et al., 2014). Relative advantage is the degree to which an individual perceives an innovation to be better than the predecessor to that innovation (Rogers, 1995). Relative advantage has been studied quite widely in areas such as electronic business (Zhu et al., 2006), internet and mobile banking (Tan andTeo, 2000), development of a web attendance, and instant messaging (Hsu et al., 2007). An Individual’s awareness of the relative advantage of a new technology such as a mobile payment service can take many forms. Although relatively broad, relative advantage operates well as a perceptual construct because it can be generalized across fields to meter perceived fulfillment of an intended aim over an earlier innovation (Moore and Benbasat,1991). In the situation of mobile banking, time saving, money and convenience have been quoted as relative advantages. In this context, relative advantage is defined as the ready to impact on the customer satisfaction. Ease of use is the degree to which the use of a personal perspective of innovation is relatively free from physical or mental effort (Moore and Benbasat, 1991). This structure has been used in a variety of technology adoption studies as adoption of mobile services (López-Nicolás et al., 2008), the virtual store and e-commerce (LD Chen et al., 2002), m-wallet systems (Amoroso and Magnier-Watanabe, 2012), and instant messaging (Hsu et al, 2007; Van Slykeet al., 2007). Due to the mobile payment service represents a fixed payment method such as credit cards, debit cards, cash and checks, it is important that consumers view the mobile payment service at least as easy to use as the previous method. This attribute refers to the level of risk in the use of an innovation (Ram and Sheth 1989). Awareness of the risks of customers often arises due to the doubt concerning the level of conflict between the judge and actual behavior, and technology does not give the expected results and the loss of the consequences (Chen 2008; Koenig-Lewis 2010; Lee et al., 2007). In technology adoption, there is research evidence of the importance of the perception of risk in deploying new technology or services (Ndubisi and Sinti 2006). In the context of mobile banking, risk awareness is even more important because of the threat to security and privacy. Second, the fear of losing the PIN code can also cause security threats (Al-Jabri and Sohail, 2012). Third, some users also feared that hackers can access their bank account via stolen PIN (Poon 2008). Finally, some users may also fear the loss or theft of mobile devices with stored data (Al-Jabri and Sohail, 2012). 438
Customer satisfaction corresponds to the assessment of whether the product or service meets the needs and expectations of that person (Wilson et al, 2012). It also has an emotional dimension, related to customer's judgment about whether the product or service has provided a satisfactory level of consumption related to consumers (Oliver 1997). The satisfaction of the customer occupies a central position in marketing research and in the organization, because it is an important factor in the patronage, loyalty and positive word of mouth behavior and the performance of firm (Oliver 1997). Day (1984) described satisfaction as a post-choice evaluative judgment concerning a specific purchase selection. The level of satisfaction depends on the difference between the received results and expectations, if actual results are less than expected, the customers are not satisfied. In opposite, if actual results correspond with the expectations, the customer will be satisfied, if fact the result is higher than expected, and very satisfied clients. Basically, the satisfaction is a feeling that the surface from a process evaluation. 2.2 Hypothesis Perceived usefulness is a key factor in recognizing the links related to information system and computing (Venkatesh and Davis, 1996) and it is same for the area of mobile commerce (Wang et al., 2006). Therefore, if the mobile banking is useful customers can be motivated to use the facilities and services very frequently and thereby improve their satisfaction level. Thus, it can be hypothesized that: H1: Perceived usefulness will positively influence customer satisfaction. When considering the technology link to the product, the ease of use of it has always been very critical issues in customer satisfaction. When comparing mobile banking with internet banking and ATM, it could be seen that the mobile handsets as a device which has a great effect though it is relatively small and complex. When a customer finds that any system is easy to study, it creates a positive impact towards the customer and it will lead to his orher satisfaction. Therefore, it can be hypothesized that: H2: Perceived ease of use will positively influence customer satisfaction. Relative advantage states that customer should be able to understand very simply the benefit of mobile banking over its predecessors. When customers are fully convinced that mobile banking services offer more benefit than the old technologies (i.e. Internet Banking, ATM, Debit and credit cards), it gives them a better customer satisfaction. The relative benefit can come in many ways for example improves the strength in economic benefit, convenience and the quality of individual image. All these lead to customer satisfaction (Rogers 1995). Thus, it can be hypothesized that: H3: Relative advantage will positively influence customer satisfaction. 439
When making the customer satisfaction there is a higher risk which plays a main role in offering services to them. When it comes to taking the risk, it is very low where the willingness to use and satisfaction is at the highest. In Taiwan, (Wu and Wang, 2015) have found that there is a very clear statistically significant relationship between the risk and the intention to use mobile commerce by customers and it is always to their satisfaction. Therefore, it can be hypothesized that: H4: Perception of riskwill positively influence customer satisfaction. 2.3. Research Methodology 2.3.1Questionnaire Design First, this study explored of the plenty of different previous literatures to build up the questionnaire items for the survey. Items were designed from both the literatures and expanded by the researcher based on actual situation in Ho Chi Minh City, Vietnam. After that, the questionnaire was designed initially by an English version. To verify conceptual equality, the survey questionnaires was translated into Vietnamese to implement the survey and to collect data and then converted backward into English. 2.3.2 Data Collection and Sampling The main goals of this survey are to obtain reliable qualitative data to interpret the features of the mobile banking in Ho Chi Minh City, Vietnam. To have the valid response rate and accuracy for this survey, the total of questionnaires was delivered online and gave by paper. This study used the questionnaire survey which conducting the requirements of process to fulfill the survey. Respondents were asked how long they used mobile banking services. The got back questionnaires were examined carefully for omissions, legibility and consistency. 3. Results 3.1 Demographic Analysis The demographic questions involve overall information such as gender, age, education, occupation, monthly income and experience. After collecting questionnaires from 250 survey respondents, among the respondents, there is a slight difference in gender. According to statistics there are 148men the rest is female statistics with 102people. In relation to age, from 21-30 years old are the majority with 151of the total 250respondents. Next, the data of aged from 31 to 40 accounted for 30.8%, with 77people. Most respondents have bachelor’s degrees accounted for 50%. The number of customers using mobile banking services was mostly 125in totally 250people who are officers. Those with incomes from 5 million to 7.5 million accounted for 18.8%. To increase the effectiveness of research, surveys based on people who already have experience using mobile banking services. According to the survey results, the experienced people use mobile banking services less than one year accounted for 36people. Those who have experience of using more than 2 years accounted for 59people participated in the survey. 440
3.2 Reliability Analysis The Cronbach’s α (alpha) coefficient and the descriptive statistics of this research were illustrated in Table 1. The Cronbach’s α was computed to examine the reliability of the questionnaire and the result of aggregate 20 items in this research is 0.848. Hair et al. (2006) demonstrated that if the Cronbach’s α is greater than 0.7 manifest it has moderately high reliability. In addition, the Cronbach’s α coefficients of aggregate 20 items are higher than 0.7 manifest. It implies that the estimation of this investigation of aggregate 20items is high reliability. <Table 1 is about here> 3.3Factor Analysis Factor analysis of this research was performed on 20 items by sequent conception of research’s fundamental model. As indicated by the outcomes illustrate in the table 2, an aggregate of 20items were arranged into four attributes with the accumulation percentage of explained variance being 61.63%. In addition, the principal components analysis proposed four components, which accounted for 61.63% of aggregate variable. The particular of factor analysis which was introduced by the Kaiser-Meyer-Olkin (KMO) test of the data sufficiency was 0.814(higher than 0.5). Bartlett’s Test of Sphericity reached statistic significant with Sig 0.000 (p<0.001). <Table 2 is about here> In the table 3 below, the first factor is most highly correlated to RA3. The second factor is most highly correlated to PU4. The next factor is most highly correlated to PR3and the last factor is most highly correlated to PEOU3. <Table 3 is about here> 3.4 Regression Analysis Correlation investigation determines the power of the relationship as well as the level of relationship between variables. The relationship exists when significant value (p) < 0.05. Table 4 below presented the results of the Pearson’s correlations of all five factors. The outcome indicates customer satisfaction has significantly positive correlations with perceived usefulness, perceived ease of use, relative advantage and perception of risk. <Table 4 is about here> In table 5, R2-value which is 0.469 means that approximately 47% of the variance in customer satisfaction can be explained by four independent factors involve perceived usefulness, perceived ease of use, relative advantage and perception of risk. The F-test value is highly significant (p<0.01), so the model in this research is consistent with the data <Table 5 is about here> 441
Regression analysis was used to examine the associated between dependent variable and independent variables. The multiple regression outcomes to test on Hypothesis 1, Hypothesis 2, Hypothesis 3, Hypothesis 4. In the table 6 above provides concern on the four independent variables which are perceived usefulness, perceived ease of use, relative advantage and perception of risk. The outcome indicated that there is a statistical significance between four factors perceived usefulness, perceived ease of use, relative advantage, perception of risk and customer satisfaction. The influence of “perceived usefulness” (β = 0.347, p<0.05) is significant and its coefficient is positive. The hypothesis 1 is supported. Similarly, the influence of “perceived ease of use” (β = 0.214, p<0.05) is significant and its coefficient is positive. The hypothesis 2 is supported. Moreover, the influence of “relative advantage” (β = 0.160, p<0.05) is significant and its coefficient is positive. The hypothesis 3 is supported. Finally, the influence of “perception of risk” (β= 0.133, p<0.05) is significant and its coefficient is positive. The hypothesis 4is supported. <Table 6 is about here> 3.5 ANOVA Analysis by Demographic Profile This study utilized ANOVA to verify the relationship between data variables at significant level of .05. In the result on the table 7, we can realize gender which a have the significant outcome. Exception perceived usefulness, gender affects positively on perceived ease of use, relative advantage and perception of risk. We can predict when male and female answer the survey questionnaire they have different answer between male and female. On the other hand, the outcome of age, education, occupation, income and experiences not significant. <Table 7 is about here> The result of research analysis via the demographic analysis, reliability analysis, factor analysis, Pearson correlation tests and regression analysis. We can realize that the independent factors affects positively the dependent factors and the hypotheses were proved by the data which was collected and analyzed clearly. The perceived usefulness is the most influential factor to perceived ease of use, relative advantage and perception of risk with the highest beta coefficient (β = 0.347). In the TAM model, perceived usefulness and perceived ease of use were demonstrated that strongly impact on customer satisfaction (Davis, 1989), hence this model is no exception. Perceived usefulness has the highest beta coefficient (β = 0.347) than perceived ease of use (β = 0.214). In addition, factor loading bigger than 0.4 that means they have practical significant. Relative advantage and perception of risk have a significant impact with β = 0.160and β = 133, respectively. In the research of Taylor and Todd (1995) point out that perceived usefulness has both direct and indirect impacted on customer satisfaction. Therefore, we can see that perceived usefulness is an important factor on this model framework. All this 442
point to the fact that all factors in model framework are suitable with this research and practical significance. 4. References Zhou, T., Lu, Y., Wang, B., (2010). Integrating TTF and UTAUT to explain mobile banking user adoption. Computers in Human Behavior, 26(4), 760-767. Mehrad, D., & Mohammadi, S. (2017). Word of Mouth impact on the adoption of mobile banking in Iran. Telematics and Informatics,34(7), 1351-1363. Dineshwar, R., & Steven, M. (2013, February). An investigation on mobile banking adoption and usage: A case study of Mauritius. In Proceedings of 3rd Asia-Pacific Business Research Conference (pp. 25-26). Laukkanen, T. (2007). Internet vs mobile banking: comparing customer value perceptions. Business process management journal, 13(6), 788-797. Shaikh, A. A., & Karjaluoto, H. (2015). Mobile banking adoption: A literature review. Telematics and Informatics, 32(1), 129-142. Hanafizadeh, P., Behboudi, M., Koshksaray, A. A., & Tabar, M. J. S. (2014). Mobile-banking adoption by Iranian bank clients. Telematics and Informatics, 31(1), 62-78. Rogers, E. M.(1995). Diffusion of Innovations (4th Eds.) ACM, The Free Press (Sept. 2001). New York, 15-23. Zhu, K., Dong, S., Xu, S. X., & Kraemer, K. L. (2006). Innovation diffusion in global contexts: determinants of post-adoption digital transformation of European companies. European journal of information systems, 15(6), 601-616. Tan, M., & Teo, T. S. (2000). Factors influencing the adoption of Internet banking. Journal of the AIS, 1(1es), 5. Hsu, C. L., Lu, H. P., & Hsu, H. H. (2007). Adoption of the mobile Internet: An empirical study of multimedia message service (MMS).Omega, 35(6), 715-726. Moore, G. C., & Benbasat, I. (1991). Development of an instrument to measure the perceptions of adopting an information technology innovation. Information systems research, 2(3), 192-222. Ram, S., & Sheth, J. N. (1989). Consumer resistance to innovations: the marketing problem and its solutions. Journal of Consumer Marketing, 6(2), 5-14. Chen, L. D. (2008). A model of consumer acceptance of mobile payment. International Journal of Mobile Communications, 6(1), 32-52. Koenig-Lewis, N.,Palmer, A., & Moll, A. (2010). Predicting young consumers' take up of mobile banking services. International journal of bank marketing, 28(5), 410-432. Lee, K. S., Lee, H. S., & Kim, S. Y (1970). Factors influencing the adoption behavior of mobile banking: a South Korean perspective. The Journal of Internet Banking and Commerce, 12(2), 1-9. Oly Ndubisi, N., & Sinti, Q. (2006). Consumer attitudes, system's characteristics and internet 443
banking adoption in Malaysia. Management Research News, 29(1/2), 16-27. Al-Jabri, I., & Sohail, M. S. (2012). Mobile banking adoption: Application of diffusion of innovation theory. Journal of Electronic Commerce Research, 13, 379-391. Poon, W. C., “Users’ adoption of e-banking services: the Malaysian perspective,” Journal of Business and Industrial Marketing, Vol. 23, No. 1: 59-69, 2008. Wilson, A., Zeithaml, V. A., Bitner, M. J., & Gremler, D. D. (2012). Services marketing: Integrating customer focus across the firm. McGraw Hill. Oliver, R. L. (1997). Emotional expression in the satisfaction response. Satisfaction: A behavioral perspective on the consumer, 291-325. Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of marketing research, 460-469. Day, R. L. (1984). Modeling choices among alternative responses to dissatisfaction, in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 496-499. Chitungo, S. K., & Munongo, S. (2013). Extending the technology acceptance model to mobile banking adoption in rural Zimbabwe. Journal of Business Administration and Education, 3(1). Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology, .MIS Quarterly, 319-340. Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology acceptance model: Four longitudinal field studies. Management Science, 46(2), 186-204. Bagozzi, R. P., Davis, F. D., & Warshaw, P. R. (1992). Development and test of a theory of technological learning and usage. Human relations, 45(7), 659-686. Amoroso, D. L., & Magnier-Watanabe, R. (2012). Building a research model for mobile wallet consumer adoption: the case of mobile Suica in Japan. Journal of theoretical and applied electronic commerce research, 7(1), 94-110. Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multivariate data analysis 6th ed., Upper Saddle River: Pearson Prentice Hall. Taylor, S., and Todd, P. A. (1995). Assessing IT usage: The role of prior experience. MIS Quarterly, 19(2), 561–570 5. Appendix Table 1. Descriptive statistics and the Cronbach’sαcoefficients (N=250, overall Cronbach’s α =0.864 Itme Labels Mean Std. Derviation Perceived usefulness (PU)(5 items, Cronbach’s α=0.772) 3.97 0.568 PU1 3.99 0.807 444
PU2 4.01 0.794 PU3 3.97 0.799 PU4 3.91 0.792 PU5 3.98 0.739 Perceived ease of use (PEOU)(4 items, Cronbach’s α=0.707) 4.23 0.525 PEOU1 4.18 0.704 PEOU2 4.24 0.697 PEOU3 4.32 0.718 PEOU4 3.21 0.760 Relative advantage (RA)(5 items, Cronbach’s α=0.0.854) 4.07 0.716 RA1 4.07 0.889 RA2 4.20 0.901 RA3 4.02 0.852 RA4 4.04 0.929 RA5 4.02 0.938 Perception of risk (PR)(3 items, Cronbach’s α=0.876) 3.90 0.992 PR1 3.93 1.094 PR2 3.97 1.031 PR3 3.80 1.195 Customer satisfaction (CS)(3 items, Cronbach’s α=0.734) 4.12 0.612 CS1 3.92 0.817 CS2 4.24 0.779 CS3 4.23 0.671 Table 2 Factor analysis Barlett’s Test Sig Number of items Accumulation Number of component KMO 0.000 20 Percentage of explained extracted 0.814 4 Variance (%) 61.63 4 Table 3. Rotated component matrix Compnent 123 RA3 0.808 RA2 0.803 RA4 0.777 RA1 0.704 RA5 0.698 445
PU4 0.777 PU1 0.733 PU5 0.704 PU2 0.703 PU3 0.666 PR3 PR2 0.900 PR1 0.861 PEOU3 0.791 PEOU1 PEOU4 0.755 PEOU2 0.697 0.682 0.653 PU Pearson correlation Table 4. Pearson correlation Sig. (2-tailed) PU PEOU RA PR CS 1 PEOU Pearson correlation Sig. (2-tailed) 0.236** 1 1 1 1 0.000 RA Pearson correlation 0.117 0.409** 0.484** 0.439** Sig. (2-tailed) 0.250 0.000 0.000 0.000 0.101 0.272** 0.531** PR Pearson correlation 0.110 0.000 0.000 Sig. (2-tailed) 0.233** 0.392** 0.000 0.000 CS Pearson correlation Sig. (2-tailed) ** p<0.01, * P<0.05 Table 5. The regression analysis hypothesis results CS R2 0.372 N 250 Adjusted R2 0.362 F 36.277 Sig.(<0.05) 0.000** Table 6. Coefficients of the regression analysis hypothesis Independent Unstandardized Standardized t value variable Coefficients Coefficients 446
Constant 1.043 0.347 3.361** PU 0.297 0.214 53661** 0.132 0.160 3.678** PEOU 0.187 2.817** RA 0.144 0.133 2.556** PR ** p<0.05 Table 7. The result of ANOVA by personal profile Factor Gender Age Education Occupation Income Experience PU 0.43** 0.610 0.564 0.736 0.143 0.343 0.031** 0.845 0.690 0.133 0.553 0.589 PEOU 0.000** 0.958 0.425 0.913 0.955 0.300 RA 0.000** 0.996 0.827 0.495 0.494 0.111 PR ** p<0.05 6. Acknowledgments and Legal Responsibility The authors would like to thank all colleagues and students who contributed to this study. However, The errors idiocies and inconsistencies remain my own. 447
Search
Read the Text Version
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
- 26
- 27
- 28
- 29
- 30
- 31
- 32
- 33
- 34
- 35
- 36
- 37
- 38
- 39
- 40
- 41
- 42
- 43
- 44
- 45
- 46
- 47
- 48
- 49
- 50
- 51
- 52
- 53
- 54
- 55
- 56
- 57
- 58
- 59
- 60
- 61
- 62
- 63
- 64
- 65
- 66
- 67
- 68
- 69
- 70
- 71
- 72
- 73
- 74
- 75
- 76
- 77
- 78
- 79
- 80
- 81
- 82
- 83
- 84
- 85
- 86
- 87
- 88
- 89
- 90
- 91
- 92
- 93
- 94
- 95
- 96
- 97
- 98
- 99
- 100
- 101
- 102
- 103
- 104
- 105
- 106
- 107
- 108
- 109
- 110
- 111
- 112
- 113
- 114
- 115
- 116
- 117
- 118
- 119
- 120
- 121
- 122
- 123
- 124
- 125
- 126
- 127
- 128
- 129
- 130
- 131
- 132
- 133
- 134
- 135
- 136
- 137
- 138
- 139
- 140
- 141
- 142
- 143
- 144
- 145
- 146
- 147
- 148
- 149
- 150
- 151
- 152
- 153
- 154
- 155
- 156
- 157
- 158
- 159
- 160
- 161
- 162
- 163
- 164
- 165
- 166
- 167
- 168
- 169
- 170
- 171
- 172
- 173
- 174
- 175
- 176
- 177
- 178
- 179
- 180
- 181
- 182
- 183
- 184
- 185
- 186
- 187
- 188
- 189
- 190
- 191
- 192
- 193
- 194
- 195
- 196
- 197
- 198
- 199
- 200
- 201
- 202
- 203
- 204
- 205
- 206
- 207
- 208
- 209
- 210
- 211
- 212
- 213
- 214
- 215
- 216
- 217
- 218
- 219
- 220
- 221
- 222
- 223
- 224
- 225
- 226
- 227
- 228
- 229
- 230
- 231
- 232
- 233
- 234
- 235
- 236
- 237
- 238
- 239
- 240
- 241
- 242
- 243
- 244
- 245
- 246
- 247
- 248
- 249
- 250
- 251
- 252
- 253
- 254
- 255
- 256
- 257
- 258
- 259
- 260
- 261
- 262
- 263
- 264
- 265
- 266
- 267
- 268
- 269
- 270
- 271
- 272
- 273
- 274
- 275
- 276
- 277
- 278
- 279
- 280
- 281
- 282
- 283
- 284
- 285
- 286
- 287
- 288
- 289
- 290
- 291
- 292
- 293
- 294
- 295
- 296
- 297
- 298
- 299
- 300
- 301
- 302
- 303
- 304
- 305
- 306
- 307
- 308
- 309
- 310
- 311
- 312
- 313
- 314
- 315
- 316
- 317
- 318
- 319
- 320
- 321
- 322
- 323
- 324
- 325
- 326
- 327
- 328
- 329
- 330
- 331
- 332
- 333
- 334
- 335
- 336
- 337
- 338
- 339
- 340
- 341
- 342
- 343
- 344
- 345
- 346
- 347
- 348
- 349
- 350
- 351
- 352
- 353
- 354
- 355
- 356
- 357
- 358
- 359
- 360
- 361
- 362
- 363
- 364
- 365
- 366
- 367
- 368
- 369
- 370
- 371
- 372
- 373
- 374
- 375
- 376
- 377
- 378
- 379
- 380
- 381
- 382
- 383
- 384
- 385
- 386
- 387
- 388
- 389
- 390
- 391
- 392
- 393
- 394
- 395
- 396
- 397
- 398
- 399
- 400
- 401
- 402
- 403
- 404
- 405
- 406
- 407
- 408
- 409
- 410
- 411
- 412
- 413
- 414
- 415
- 416
- 417
- 418
- 419
- 420
- 421
- 422
- 423
- 424
- 425
- 426
- 427
- 428
- 429
- 430
- 431
- 432
- 433
- 434
- 435
- 436
- 437
- 438
- 439
- 440
- 441
- 442
- 443
- 444
- 445
- 446
- 447
- 448
- 449
- 450
- 451
- 452
- 453
- 454
- 455
- 456
- 457
- 458
- 459
- 460
- 461
- 462
- 463
- 464
- 465
- 466
- 467
- 468
- 469
- 470
- 471
- 472
- 473
- 474
- 475
- 476
- 477
- 478
- 479
- 480
- 481
- 482
- 483
- 484
- 1 - 50
- 51 - 100
- 101 - 150
- 151 - 200
- 201 - 250
- 251 - 300
- 301 - 350
- 351 - 400
- 401 - 450
- 451 - 484
Pages: