In reliability studies, the goal is to gain knowledge about a product's failure times or life expectancy. Precedence tests do not require large sample sizes and are used in reliability studies to compare the life-time distributions from two samples. Precedence tests are useful since they provide reliable results early in a life-test and the surviving units can be used in other tests. Ng and Balakrishnan (2010) proposed a precedence-type test based on the Kaplan-Meier estimator of the cumulative distribution function. A precedence-type test based on the Nelson-Aalen estimator of the cumulative hazard function has been proposed. This test was developed for both Type-II right censoring and progressive Type-II right censoring. Numerical results, including illustrative examples, critical values and a power study have been provided. The results from this test were compared with those from the test based on the Kaplan-Meier estimator.

This study investigates the use of a Web Based Learning environment to support a newly introduced Problem Based Learning curriculum at the Nelson R. Mandela School of Medicine. Questionnaires, observations and interviews form the basis of this qualitative study amongst first and second year students. Attempts are made to interrogate the way in which staff members use the system to support a constructivist learning environment. In addition similar prior research at this institution is taken into account and reported in the literature review. While not disputing the findings of earlier research the qualitative methodology used here shows some discrepancies with previous research. In particular, although learners are able to, and do, make use of the system it is predominantly used for the dissemination and retrieval of information. The basic premises of construction of knowledge are not facilitated by the Medical School's use of the Web Based Learning environment and of great concern was the small number of staff members who saw it as an integral part of the new curriculum and learning process. / Thesis (M.A.)-University of KwaZulu-Natal, 2006.

Only seven years more than one hundred years ago, 1838, Kentucky’s public school system was established. Few Kentuckians conceived of a state-wide public school system. Many of the citizens were conservative and inclined to look upon a scheme of public education with doubt and disfavor; they were accustomed to the traditional plan of private education. In Nelson County, as in the rest of Kentucky, the early sponsors of public education were confronted with the problem of building sentiment for such a public service. Their task was one of establishing schools. Their chief goal was one of providing a school service, meager as it was, within reasonable reach of each child. Although skepticism and conservatism retarded the early growth of public schools, it is now evident that a system of public elementary and secondary education meets the approval of the citizenry of Nelson County. In recent years, educators have striven to check this growth in the number of schools and school districts. The present task is one of improving schools rather than establishing schools; one of broadening educational services to meet present-day demands; one of equalizing educational opportunities in order that all pupils may be equipped for more abundant living. Whenever possible the school authorities in Nelson County have reduced the number of one-room rural schools in a consolidation program. This study gives evidence in justification of these changes.

本研究以新台幣兌美元之匯率日資料作為主要研究標的，同時加入台灣加權股價指數及金融業隔夜拆借利率之日資料作為股價與利率之代理變數，利用Beveridge-Nelson分解趨勢的方法將變數資料拆解成趨勢項與循環項之時間序列資料，藉此捕捉匯率資料具有景氣循環的特性。在循環項的序列資料，以向量自我迴歸模型來分析並予以估計，趨勢項的部分，利用共整合檢定來探討趨勢項變數間長期的均衡關係，再以向量誤差修正模型予以估計，得到未來30天期之匯率走勢。接著，再以RMSE與MAE指標來衡量不同模型之匯率預測績效，以期能找出最適之匯率預測模型。 實證研究結果發現，將匯率資料先透過Beveridge-Nelson分解趨勢的方法予以拆解後，再利用時間序列模型進行分析及預測，時間序列模型的預測能力都比原始匯率利用時間序列模型進行預測或透過ARIMA模型進行預測還要來的好。因此，根據實證研究的結果，若企業與政府在進行匯率預測的分析時，能夠考慮先將匯率資料透過Beveridge-Nelson分解方法予以處理，便能更有效提升模型的預測能力，除了企業能夠降低避險成本來提高公司整體績效，對於國家而言，有效的掌握匯率的趨勢便能夠迅速且正確的制定政策，提升國家的經濟發展。

In reliability studies, the goal is to gain knowledge about a product's failure times or life expectancy. Precedence tests do not require large sample sizes and are used in reliability studies to compare the life-time distributions from two samples. Precedence tests are useful since they provide reliable results early in a life-test and the surviving units can be used in other tests. Ng and Balakrishnan (2010) proposed a precedence-type test based on the Kaplan-Meier estimator of the cumulative distribution function. A precedence-type test based on the Nelson-Aalen estimator of the cumulative hazard function has been proposed. This test was developed for both Type-II right censoring and progressive Type-II right censoring. Numerical results, including illustrative examples, critical values and a power study have been provided. The results from this test were compared with those from the test based on the Kaplan-Meier estimator.

The relation between place attachment and pro-environmental behaviour has not been thoroughly examined. Of the few studies to have investigated this, findings conflict (Uzzell, Pol, & Badenas, 2002; Vaske & Kobrin, 2001). Possibly, these inconsistencies relate to the definition of place attachment. The current study distinguished two dimensions of place attachment: social-symbolic (civic), and physical-natural. Data were collected from 104 community members in two proximate towns. Participants indicated their levels of natural and civic attachment, evaluated the current and future local environmental quality, and reported their pro-environmental behaviours. Greater place attachment was associated with more positive current evaluations in Trail. Surprisingly, environmental evaluations did not predict pro-environmental behaviour in either town. Both types of place attachment predicted pro-environmental behaviour in Nelson, but in Trail, only natural attachment was significant. Thus, for different cities, different types of place attachment will predict pro-environmental behaviour. Implications and directions for further research are discussed.

Thesis (M.B.A. (Rhodes Investec Business School)) - Rhodes University, 2007. / A thesis submitted in partial fulfillment of the requirements for the degree of Masters in Business Administration (MBA)

Thesis (Ph. D.)--University of Missouri-Columbia, 1997. / Typescript. Vita. Includes bibliographical references (leaves 1459-1564). Also available on the Internet.

Thesis (Ph. D.)--University of Missouri-Columbia, 1997. / Typescript. Vita. Includes bibliographical references (leaves 1459-1564). Also available on the Internet.

The relation between place attachment and pro-environmental behaviour has not been thoroughly examined. Of the few studies to have investigated this, findings conflict (Uzzell, Pol, & Badenas, 2002; Vaske & Kobrin, 2001). Possibly, these inconsistencies relate to the definition of place attachment. The current study distinguished two dimensions of place attachment: social-symbolic (civic), and physical-natural. Data were collected from 104 community members in two proximate towns. Participants indicated their levels of natural and civic attachment, evaluated the current and future local environmental quality, and reported their pro-environmental behaviours. Greater place attachment was associated with more positive current evaluations in Trail. Surprisingly, environmental evaluations did not predict pro-environmental behaviour in either town. Both types of place attachment predicted pro-environmental behaviour in Nelson, but in Trail, only natural attachment was significant. Thus, for different cities, different types of place attachment will predict pro-environmental behaviour. Implications and directions for further research are discussed.