本研究的目的在暸解教師流動的型態與調動的因素，從83學年度臺灣省國民中學調動教師中抽取384人，並同時抽取380位未調動教師作為參照。 　　本研究採用虛擬依變項的迴歸分析，探討導致教師調動的因素。結果發現教師調動的型態存在往都會地區流動的現象。調動與非調動教師的比較中，在性別、教學年資、婚姻狀況與配偶的勞動參與率、自用住宅的擁有、調動次數、學校規模、以及是否為居民的變項上有顯著的不同。 　　在調動機率的影響方面，性別、教學年資、婚姻狀況與配偶的勞動參與率、調動次數、學校規模、以及薪資與福利措施的滿意度上，有著顯著的預測力。研究的結論是教育人力資源分配仍不平均，從影響教師調動機率的因素中，可以提供導引教育人力資源分配更加均衡的線索。 / The main purpose of this study is to inquire into the style about the teachers' mobility and the factors affecting such mobility. Sampling from 1994 academic year in Taiwan, the researcher focuses on the 384 cases of all mobilized teachers and their counterpart-380 nonmobilized cases. 　　A dummy dependent regression analysis is conducted to explore the factors. As a result, the phenomenon for mobility tends to flow into the metropolis. Factors affecting the mobility exhibits significant differences compared to nonmobility; such as factors among sex, seniority, marriage, inhibition, frequcncy of mobility and the scale of the school. 　　In additions, the propensity to mobile, as this study finds, has to do with their sex, seniority, marriage and the labor participation of spouse, frequcency of mobility, degree of the content about their wages and welfare, and the scale of the school. The conclusions made here are: the human resources in education is still scanty. From the factors influenced mobility probability, it can provide some clues to make the human resources in education more equal.

The main purpose of this study is to inquire into the style about the teachers' mobility and the factors affecting such mobility. Sampling from 1994 academic year in Taiwan, the researcher focuses on the 384 cases of all mobilized teachers and their counterpart-380 nonmobilized cases. A dummy dependent regression analysis is conducted to explore the factors. As a result, the phenomenon for mobility tends to flow into the metropolis. Factors affecting the mobility exhibits significant differences compared to nonmobility; such as factors among sex, seniority, marriage, inhibition, frequcncy of mobility and the scale of the school. In additions, the propensity to mobile, as this study finds, has to do with their sex, seniority, marriage and the labor participation of spouse, frequcency of mobility, degree of the content about their wages and welfare, and the scale of the school. The conclusions made here are: the human resources in education is still scanty. From the factors influenced mobility probability, it can provide some clues to make the human resources in education more equal.

No description available.

本文使用具有事前訊息的領先指標與期間利差作為預測變數,根據不同利差與落後期選擇的 Probit 模型,利用遞迴的方式預測景氣轉折點發生機率,並進一步將個別預測結果進行組合,試圖找出能降低不確定性且優於個別預測結果的方法。實證結果發現,使用 Diebold and Mariano 檢定的預測包容法為其中最優的組合方法,無論是轉折點訊號或預測誤差都能優於半數以上的個別預測。此外,本文亦估計即期景氣轉折點的發生機率,根據模型的估計結果推斷,自 2012 年 2 月至 2015 年 3 月為止,景氣仍處於擴張階段。

nÃo hà / Segundo Jensen (1968), a indÃstria de fundos mÃtuos de investimento, cuja expansÃo està prevista teoricamente pelo Teorema da SeparaÃÃo enunciado em Sharpe (1964), teria limitaÃÃes no sentido de bater o mercado em termos de performance risco-retorno mensurada pelo alfa de Jensen. Nesta ampla discussÃo, esta dissertaÃÃo se posiciona em sugerir um exercÃcio empÃrico aplicado a um cross-section contendo 243 fundos de investimentos em aÃÃes, categoria Ibovespa Ativo, o qual visa identificar que variÃveis financeiras, contÃbeis e administrativas se mostram capazes de prever no ano seguinte o sinal e a significÃncia do alfa de Jensen. Foram extraÃdos retornos diÃrios para todos os fundos nos anos de 2011 e 2012, e calculadas mÃtricas clÃssicas de retorno, risco e performance, bem como os 24 balancetes mensais e informaÃÃes administrativas do perÃodo em questÃo. Metodologicamente, as variÃveis explicativas consistem em estatÃsticas descritivas obtidas a partir de dados financeiros diÃrios e contÃbeis mensais, enquanto as performances a serem modeladas sÃo estimadas por meio do Capital Asset Pricing Model (CAPM). Dessa maneira, foi possÃvel ordenar os fundos em trÃs grupos, composto por Loosers, Draw e Winners, de acordo com suas performances em relaÃÃo ao Ãndice Ibovespa. Sendo assim, foi identificado que apenas 71 dos fundos foram capazes de performar melhor que o Ãndice Ibovespa durante o ano de 2012. Os resultados obtidos com a estimaÃÃo do arcabouÃo de Probit ordenado sugerem que fundos com maiores performances mensuradas pelos alfa de Jensen e Ãndices de Calmar e Sortino, associados a menores taxas de administraÃÃo, tendem a bater o mercado no ano seguinte. Entretanto, mÃtricas clÃssicas como desvio-padrÃo, taxa de performance e Ãndice de Sharpe (1964) nÃo se mostraram significantes. O modelo sugere, tambÃm, que a variÃvel Drawdown seja apresentada como mÃtrica eficiente de mensuraÃÃo de risco. / According to Jensen (1968), the mutual funds industry expansion is theoretically predicted by the Separation Theorem stated by Sharpe (1964), however with limitations in order to exceed the market in terms of risk-return performance measured by Jensen's alpha. In this broad discussion, this dissertation suggest an empirical exercise applied to a cross-section containing 243 stock funds, within the Ibovespa Active category, which aims to identify which financial, accounting and administrative variables are capable to predict the next year's value and the significance of the Jensen's alpha. Daily returns were extracted for all funds in 2011 and 2012, and were calculated classic metrics such as return, risk and performance. There were also extracted 24 monthly accounting balances and administrative informations for the period in question. Methodologically, the explanatory variables consist of descriptive statistics obtained from daily financial data and monthly accounting data, while the performances to be modeled are estimated using the Capital Asset Pricing Model (CAPM). Using this technic it was possible divide the funds into three groups, consisting of Loosers, Draw and Winners, according to their performances in relation to the Ibovespa index. Thus, it was discovered that only 71 funds were able to perform better than the Ibovespa Index during the year 2012. The estimation results of the ordered probit framework suggests that funds with higher performances measured by the Jensen's Alpha and with higher Sortino and Calmar ratios, associated with lower management fees tend to surpass the market in the next year. However, classical metrics like standard deviation, performance fees and Sharpe ratio (1964) were not significant. The model also suggests that the drawdown variable should be used as an efficient risk metric.

L'objet de cette thèse est d'étudier les liens qui existent entre les déséquilibres macroéconomiques et les crises, et de voir dans quelles mesures leur prise en compte peut aider une meilleure gestion des crises dans les pays de la zone euro. Les différents chapitres de cette thèse tentent d'apporter des réponses à trois questions importantes : Quels sont les indicateurs macro-financiers qui pourraient aider à mieux anticiper les épisodes de stress budgétaire dans les pays de la zone euro ? Quelles seraient les conséquences de la mise en place d'un système de garantie des dépôts bancaires sur les variables macroéconomiques et sur le comportement des investisseurs, investisseurs qui tiendraient compte du risque de défaut souverain ? Dans quelle mesure une meilleure qualité des institutions, de la gouvernance pourrait-elle aider à améliorer la croissance de long terme d'une économie contrainte sur le marché international des capitaux ? En retenant une approche de court terme, les deux premiers chapitres montrent l'importance de la confiance des marchés dans l'analyse du lien entre déséquilibres macroéconomiques et crises. Dans le troisième chapitre, nous adoptons une perspective de plus long terme pour analyser les effets de cette confiance des marchés sur la dynamique de la croissance. Notre approche est à la fois théorique et empirique. L'approche théorique se base sur les modèles DSGE (modèles d'équilibre général stochastiques dynamiques et la modélisation d'une crise dans une petite économie ouverte. L'approche empirique se focalise sur les modèles Probit/Logit sur données de panel et sur un modèle d'alerte fondé sur des signaux avancés (early warning indicators). / This dissertation consists of three essays on how macro-financial imbalances precede crises and to what extent their consideration can help better management of crises in the Eurozone countries. The different chapters of this thesis, try to answer three important questions : What are the macro-financial imbalances that exposed the Euro area countries to fiscal stress before the outbreak of the debt crises in Europe? What are the impacts of sovereign default and deposit guarantee on macroeconomic variables and on the behavior of investors ? To what extent could better institutions/governance help to improve the long-term growth in a constrained economy on the international capital market ? Using a short-term approach, the first two chapters show the importance of market confidence in analysis of the link between macroeconomic imbalances and crises. In the third chapter, we adopt a long-term perspective to analyze the effects of this market confidence on the dynamics of growth. Our approach is both theoretical and empirical. The theoretical approach is based on the DSGE models (dynamic stochastic general equilibrium models) and the modeling of a crisis in a small open economy (SOE). The empirical approach focuses on Probit/Logit models for panel data and on Signal model based on early warning indicators.

本研究預測匯率危機的方法主要是用二元分量迴歸(Binary Regression Quantiles)，此理論基礎與預測方式是使用美國學者Kords (2004)的方法，將分量迴歸運用在應變數為二元的屬質變數上之計量方法。在匯率危機計量模型中，最常使用的模型是Logit模型和Probit模型所做的分析，因此本÷究除了使用二元分量模型外，將在套入Logit模型和Probit模型，並將這三種模型加以比較，且探討匯率危機發生的原因並建立預警變數。而研究資料為十七個發展中的國家，研究時間為1981~2004年。 本研究發現由Logit模型和Probit模型中，兩模型的所預測匯率危機指標大都一致，包括有進口比例、GDP成長率、銀行外債/GDP。而且發現由二元分量回歸模型中，匯率危機預警指標有出口/GDP、貿易條件、海外直接投資/GDP、國際熱錢流入/GDP、銀行存款、GDP成長率、貪污指數，短期外債/全部外債。

臺灣原住民是這塊土地上重要的大自然文化資產之一；他們縱橫在這個島嶼的山林與海洋間千百年，生活智慧是採取與山林、海洋、大自然互動和諧的自然演替，各族群發展出自己的「原生知識」、獵場及祖靈。原住民沒有文字記載，千百年來當地族群不斷口語相傳，傳遞著各部族的生活智慧與傳承，因而孕育出各族群繽紛多彩俱獨特性的原住民文化。 自17世紀後的臺灣經歷西、荷、明鄭、滿清、日據、國民政府等四百多年不同政權的移轉統治，就不斷受到外族武力、政治、社會、經濟、文化等政策性的壓迫，代表當地族群生活智慧的原生知識被視為無商業價值而棄置；各級政府在推動政策時並未對其文化的差異性而有特殊考量，輔導其達到自主發展的目標，致使他們無法融入主流文化，造成所有原住民族不但在傳統與現代銜接上出現結構上的問題，加上原住民對金錢的儲蓄投資觀念薄弱，他們都成為經濟弱勢族群。 原住民家庭經濟狀況低落為長期以來存在的問題， 本研究旨在探討原住民投資收益佔經濟戶長收入的比例。除應用行政院原民會95年臺灣原住民經濟狀況調查報告之數據資料作實證分析外，亦蒐集國內外文獻、論文資料予以分析，藉以由原住民投資與儲蓄之狀況，瞭解臺灣原住民所得偏低，投資與儲蓄的習慣尚待建立，是否就是其成為經濟弱勢的主因，並期待透過本研究找出原住民投資的決定因素，藉以找出解決方案，作為政府擬定原住民政策時的參考，以幫助原住民走出貧窮及悲情。 經本研究之主要發現：解釋變數儲蓄、性別、婚姻狀況、教育程度、年齡、每月租金、房貸、各項消費信貸、個人所得及家庭所得等，經以probit與logit之模型作迴歸分析的結果，都是具有影響力。以研究結果對政府部門的建議為，一要提升原住民教育程度；二為提高勞保就保率與降低職災；三為加強宣導戒除菸酒及檳榔習慣；四為原住民應有儲蓄及投資習慣；五為原住民本身也應走出悲情，不再藉酒澆愁，重新展現他們應有的自信與能力之優越感。

This study documented the psychometric development of the Adaptive Leadership Competency Profile (ALCP). The ALCP was derived from a qualitative database from the National Science Foundation project (NSF 9422368) and the academic body of literature. Test items were operationalized, and subject matter experts validated 11 macro-leadership competencies and 65 items. Rasch rating scale measurement models were applied to answer the following questions: (a) How well do the respective items of the ALCP fit the Rasch rating scale measurement model for the 11 scales of the ACLP? (b) How well do the person's abilities fit the Rasch rating scale measurement model, using the 11 scales of the ALCP? (c) What are the item separation and reliability coefficients for the 11 ALCP scales? (d) What are the person separation and reliability coefficients for the 11 ALCP scales? This study also sought to discern whether the ALCP could predict leader effectiveness as measured by the likelihood ratio index and frequency of correct predictions indices. The WINSTEPS and LIMDEP programs were used to obtain Rasch calibrations and probit estimates, respectively. The ALCP profiles the frequency and intensity of leadership behavior. Composite measures were calculated and used to predict leadership effectiveness. Results from this study validated 10 competencies and 55 items.

A common problem in regression analysis is that some covariates are measured with errors. In this dissertation we present simulation-based approach to estimation in two popular regression models with a categorical response variable and classical measurement errors in covariates. The first model is the regression model with a binary response variable. The second one is the proportional odds regression with an ordinal response variable. In both regression models we consider method of moments estimators for therein unknown parameters that are defined via minimizing respective objective functions. The later functions involve multiple integrals and make obtaining of such estimators unfeasible. To overcome this computational difficulty, we propose Simulation-Based Estimators (SBE). This method does not require parametric assumptions for the distributions of the unobserved covariates and error components. We prove consistency and asymptotic normality of the proposed SBE's under some regularity conditions. We also examine the performance of the SBE's in finite-sample situations through simulation studies and two real data sets: the data set from the AIDS Clinical Trial Group (ACTG175) study for our logistic and probit regression models and one from the Adult Literacy and Life Skills (ALL) Survey for our regression model with the ordinal response variable and mismeasured covariates. / October 2016