Characteristics of fuzzy synthetic decision methods for measuring student achievement

Traditional method of student achievement evaluation only use arithmetic mean and convert them to rankings, but this does not provide further explanatory information to proceed with more reasonable evaluations, decisions, and interpretations for the learning achievements of students, and provide a fair and appropriate consideration of the evaluation results. Therefore, this study attempts to introduce four types of fuzzy synthetic decision methods in actual scores for evaluating and ranking student’s academic achievement. Using the synthetic decision method of fuzzy theory, the four types of composite operations are used in conjunction with the membership function, and finally fuzzy means are used to express the diverse evaluation results of students. This study uses junior high school first year students in central Taiwan as research subjects, selecting the actual mid-term exam results of the gifted class and general class as research data. The fuzzy synthetic decision method is applied through four types of composite operations, proposing a ranking system that is more diverse and precise than the traditional average method. Finally, this study proposes the characteristics of the four types of fuzzy composite operations, considering the most suitable composite operations for classes with different characteristics. Results of this study can be used as a reference for educators in the field and future researchers. The contribution of this study is to provide the fuzzy grade calculation methods that are suited to students with different characteristics, in order to achieve diverse and precise ranking evaluations.     

IS FULL EFFICIENCY ACHIEVABLE? AN EMPIRICAL STUDY OF THE BANKING INDUSTRY IN TAIWAN

This paper aims to propose a feasible framework for estimating the efficiency of input factors for banks and the adjustment of these factors necessary to achieve full efficiency in the short and long term. Based on a sample of 39 Taiwanese banks during 1999–2011, the framework recommends a scheme along with a set of adjustments empirically that allow the banks to not only achieve full efficiency but also save 10.3% of the total costs in the short term and 8.8% in the long term. This scheme amounts to short‐term and long‐term savings of NT$3.6 billion and NT$3.1 billion, respectively. (JEL C23, D24, G21)     

Real Earnings Management Uncertainty and Corporate Credit Risk

Abstract

This study examines the accounting information uncertainty effects on corporate credit risk from the perspective of real earnings management (RM) activities by investigating 9565 American bond observations from year 2001 to 2008. The main results show that the volatilities of RM activities significantly and positively affect corporate bond yield spreads when well-known bond spread determinant variables are controlled. In addition, the results are robust to alternative model specifications, including the suspect firm analyses, another less ambiguous measure of abnormal cash flows from operations, and abnormal production cost analyses in manufacturing industry or with control of the input price variation. This research also finds that the positive effects of RM volatilities become weaker if a firm has a lower credit rating. Finally, our results remain hold with considering endogeneity issues and analyst characteristic variables and for another estimation period of RM volatilities.     

LABOUR MARKET ACTIVITY OF FOREIGN SPOUSES IN TAIWAN: EMPLOYMENT STATUS AND CHOICE OF EMPLOYMENT SECTOR

The present study examines the employment status and choice of employment sector of female foreign spouses from Southeast Asia and Mainland China in Taiwan. The conceptual framework is based on the family labour supply model, human and social capital theory, and immigrant assimilation theory. Our findings indicate that in regard to employment status, family background variables, including the presence of small children and husbands' characteristics, play a more significant role in determining the employment probability for these foreign spouses than do human capital variables. In particular, for spouses from Southeast Asia, each additional child is correlated with a decrease in working probability of 11.3%, whereas college education has an insignificant effect on their employment probability. Employment assimilation for these marriage immigrants may be confirmed by the finding that the employment probability of foreign spouses rises rapidly with the number of years that have elapsed since migration. As for the choice of employment sector, a strong linkage between the employment sector of the foreign spouses and their husbands' employment sector is found in this study.     

Personalized mobile advertising: Its key attributes,trends, and social impact

Advertising media are a means of communication that creates different marketing and communication results among consumers. Over the years, newspaper, magazine, TV, and radio have provided a one-way media where information is broadcast and communicated. Due to the widespread application of the Internet, advertising has entered into an interactive communications mode. In the advent of 3G broadband mobile communication systems and smartphone devices, consumers' preferences can be pre-identified and advertising messages can therefore be delivered to consumers in a multimedia format at the right time and at the right place with the right message. In light of this new advertisement possibility, designing personalized mobile advertising to meet consumers' needs becomes an important issue. This research uses the fuzzy Delphi method to identify the key personalized attributes in a personalized mobile advertising message for different products. Results of the study identify six important design attributes for personalized advertisements: price, preference, promotion, interest, brand, and type of mobile device. As personalized mobile advertising becomes more integrated in people's daily activities, its pros and cons and social impact are also discussed. The research result can serve as a guideline for the key parties in mobile marketing industry to facilitate the development of the industry and ensure that advertising resources are properly used.     

Value relevance of financial statements in convergence with IFRS: analyses in the abnormal pricing error method

We use an econometrically distinct method to test whether the value relevance of financial statements, especially during convergence with International Financial Reporting Standards (IFRS), has improved in Taiwanese firms in 1990–2011. In contrast to a prior study, our results show that given transformation between two high-quality accounting standards, convergence with IFRS will not lead to further increases in value relevance of financial statements. Our findings may provide implications for international standard-setters and regulators, especially countries undecided about adopting or converging with IFRS.     

Is there the threshold effect of cigarette price on smoking prevalence? A cross-country panel data analysis

This article shows new cross-country evidences by empirically investigating the joint effects of cigarette price levels and joining the World Health Organization Framework Convention on Tobacco Control (WHO FCTC) on smoking prevalence in 74 countries over the period of 2000, 2003, 2005 and 2006. We assessed cigarette price elasticity for three national income levels using different databases on cigarette price from the Economist Intelligence Unit (EIU), demographic and socioeconomic country characteristics from the World Bank and adjusted smoking prevalence rates published by various yearly WHO reports on the global tobacco epidemic from 2000 to 2010. A panel threshold regression was used to capture the nonlinear effects that cigarette prices on smoking prevalence at the three national income levels endogenously determined by estimation. Our findings supported the evidence that joining the WHO FCTC would have a positive effect on reducing cross-country smoking prevalence, especially among countries with low- and medium-income levels. Moreover, some simulated results show that a price hike of 10% would reduce smoking prevalence in countries with national income levels equal to or less than US$1900 and by 7.2% in countries with national income levels between US$1900 and US$2510 more than those with national income levels that are higher than US$2510.     

Epidemic forecasting with a new fuzzy regression equation

The traditional fuzzy regression model involves two solving processes. First, the extension principle is used to derive the membership function of extrapolated values, and then, attempts are made to include every collected value with a membership degree of at least h in the fuzzy regression interval. However, the membership function of extrapolated values is sometimes highly complex, and it is difficult to determine the h value, i.e., the degree of fit between the input values and the extrapolative fuzzy output values, when the information obtained from the collected data is insufficient. To solve this problem, we proposed a simplified fuzzy regression equation based on Carlsson and Fullér’s possibilistic mean and variance method and used it for modeling the constraints and objective function of a fuzzy regression model without determining the membership function of extrapolative values and the value of h. Finally, we demonstrated the application of our model in forecasting pneumonia mortality. Thus, we verified the effectiveness of the proposed model and confirmed the potential benefits of our approach, in which the forecasting error is very small.     

A final price prediction model for english auctions: a neuro-fuzzy approach

Online auction has now been a popular mechanism in setting prices for internet users. However, auction price prediction, involving the modeling of uncertainty regarding the bidding process, is a challenging task primarily due to the variety of factors changing in auction settings. Even if all the factors were accounted for, there still exist uncertainties in human behavior when bidding in auctions. In this paper, three models, regression, neural networks and neuro-fuzzy, are constructed to predict the final prices of English auctions, using real-world online auction data collected from Yahoo-Kimo Auction. The empirical results show that the neuro fuzzy system can catch the complicated relationship among the variables accurately much better than the others, which is of great help for the buyers to avoid overpricing and for the sellers to facilitate the auction. Besides, the knowledge base obtained from neuro fuzzy provides the elaborative relationship among the variables, which can be further tested for theory building.