A new random permutation test in ANOVA models

This paper developed an exact method of random permutations when testing both interaction and main effects in the two-way ANOVA model. The method of this paper can be regarded as a much improved model when compared with those of the previous studies such as Still and White (1981) and ter Braak (1992). We further conducted a simulation experiment in order to check the statistical performance of the proposed method. The proposed method works relatively well for small sample sizes compare with the existing methods. This work was supported by Korea Science and Engineering Foundation Grant (R14-2003-002-0100)     

Association of educational level and child sex ratio in rural and urban India

Utilizing data from the Census of India, this study compared child sex ratio in rural and urban regions of India and analyzed whether the child sex ratio was associated with mother’s education level. The child sex ratios in the rural and urban regions throughout India were analyzed using the two-sample and paired Student’s t-test. Further, the Chi-square test for trend in binomial proportions was used to assess the association between child sex ratios and mother’s education levels in rural and urban areas. Analysis of the data showed that there was a significant difference between the child sex ratios in rural areas versus urban areas. In addition, the Chi-square test for trend showed that there was a significant association of the child sex ratio and the educational levels of their mothers. This trend occurred in both rural and urban areas.     

Extreme Copulas and the Comparison of Ordered Lists

We introduce two extreme methods to pairwisely compare ordered lists of the same length, viz. the comonotonic and the countermonotonic comparison method, and show that these methods are, respectively, related to the copula T _M (the minimum operator) and the Ł ukasiewicz copula T _L used to join marginal cumulative distribution functions into bivariate cumulative distribution functions. Given a collection of ordered lists of the same length, we generate by means of T _M and T _L two probabilistic relations Q ^M and Q ^L and identify their type of transitivity. Finally, it is shown that any probabilistic relation with rational elements on a 3-dimensional space of alternatives which possesses one of these types of transitivity, can be generated by three ordered lists and at least one of the two extreme comparison methods.     

Transparency in the European Bond Market

The European bond market is made up of many segments characterised by varying levels of liquidity and of pre- and post-trade transparency. The possibility of extending the transparency provisions envisioned in the Directive 2004/39/EC, disciplining the functioning of markets in financial instruments in Europe (MiFID), from the equity to the bond market is currently being discussed. This poses risks to the status quo through trading fragmentation and strong negotiating power of a few large dealers, particularly with respect to small retail investors. The main conclusion of the paper is that if MiFID transparency rules are to be extended to the bond market, this must be “real” transparency, meaning pre-trade dissemination of real-time executable prices and a transparent order book. Under current provisions, this kind of information is available if not directly to everybody, at least directly to dealers and indirectly to professional and retail clients through business-to-business electronic platforms.     

Online coupon consumption problem

Nowadays, it is popular that the dealer makes profits by selling a kind of discount coupons, which can be used as money to purchase commodities with total cost less than or equal to the face value of the coupon. We can purchase a coupon at a price of 0<s≤1 times its face value and the number of potential purchasable coupons is a given integer l. The customer has the option to buy the goods by cash completely or by a discount coupon. However, each piece of goods can only use one coupon and the coupon used must have enough balance for the goods. The objective is to minimize the total cost for purchasing all the goods. In this paper, we reduce the problem to a special bin packing model. We consider the online problems for all 0<s≤1 and 1≤l≤∞. We present optimal online algorithms for all 0<s≤1 when l=∞ and l=1. For 2≤l<∞, we give both a lower bound and an algorithm, and show the algorithm is optimal for l=2. A preliminary version of this paper appeared in the proceedings of the 1st International Symposium on Combinatorics, Algorithms, Probabilistic and Experimental Methodologies, Lecture Notes in Computer Science. Y. Jiang supported by Natural Science Foundation of Zhejiang Province (Y605316). Z. Tan supported by Natural Science Foundation of China (10671177, 60021201).     

Estimating the Relative Success of Local Authorities at Labour-Market Integration of Immigrants

We propose a method for estimating indicators of the success of local authorities at integrating immigrants in the labour-market. The proportion of time in employment of individuals is used to measure labour-market integration, and we correct for differences in characteristics of immigrants and local labour markets using a two-limit Tobit model. This method is applied to a large administrative micro dataset which contains information on all immigrants in Denmark. We compare the estimated indicators of integration to corresponding uncorrected measures, and conclude that the corrections have a significant effect on the ranking of local authorities.     

Einsatz von Simulationspatienten im Kommunikations- und Interaktionstraining für Medizinerinnen und Mediziner (Medi-KIT): Bedarfsanalyse — Training — Perspektiven

Communication skills are of central importance within daily medical life providing effective treatment. In the modern high-tech field of medicine, trustful conversations play a prominent role. In order to help students of the Heidelberg Medical Faculty to be ready for their highly responsible role as communication partner, training is carried out in the framework of HeiCuMed (short for Heidelberger Curriculum Medicinale) using specially trained simulation patients. The communication and interaction training program (referred to as Medi-KIT) enables to practice conducting conversations with challenging communication partners and grave conversational content in difficult situations. Feedback from the standardized patients, as well as from peers and tutors represents a central didactic element.In order to secure knowledge transfer into daily clinic life and as well to build a relation to cognitive knowledge transfer, all medical students are assigned, alongside Medi-KIT lessons, to a ward, and training units are coordinated in terms of content to correspond with symptom-oriented lectures and parallel POL-groups (problem-oriented learning groups). In addition, communication learning objectives are examined at the end of the semester in the form of a clinical practical exam (OSCE: Objective Structured Clinical Examination) and at patient’s bedside (Mini-CEX: Mini Clinical Examination). The goal in imparting communication skills within medical education is that the longitudinal communication curriculum be maintained and continued in subsequent medical specialization studies.     

Connectivity in urbanizing landscapes: The importance of habitat configuration,urban area size,and dispersal

Human activities affect both the amount and configuration of habitat. These changes have important ecological implications that can be measured as changes in landscape connectivity. I investigated how urbanization interacts with the initial amount and aggregation of habitat to change dispersal potential, restoration potential, and the risk of spatially extensive disturbances. I used a factorial set of simulated landscapes and subjected each landscape to habitat loss by overlaying 66 different US urban areas. I used a common connectivity metric, CONNECT, to assess the magnitude and direction of changes for a range of dispersal distances. My results show that the relationship between habitat loss and connectivity loss is non-linear and subject to interactions between the spatial patterns of habitat distribution, urban morphology, and dispersal capabilities. The implications of a given urban form vary widely as a function of habitat distribution and dispersal capabilities. This implies that impact assessments, restoration activities, and conservation planning should consider historical habitat distribution when evaluating observed changes in connectivity. While my results clearly show that more aggregated or continuous habitats are more vulnerable to connectivity loss, this approach can also be used to identify landscapes where restoring connectivity will be particularly effective, for example through placement of stepping stone habitats.     

Differences in Delaying Motherhood Across European Countries: Empirical Evidence from the ECHP

Motherhood postponement and fertility decline have been observed in almost all developed countries. In this article, we aim to assess some of the potential determinants of first childbirth timing in Europe, in a comparative perspective, using data from the European Community Household Panel Survey (ECHP). We estimate, separately by country, hazard rates for the timing at first childbirth conditioning to education and work characteristics. Moreover, in order to explain differences between European countries, we decompose the differences between rates in the part due to the national population composition by specific characteristics and the part due to different propensities for women with given characteristics.