It is widely considered that the raise of carbon dioxide is a main causeof global warming, but nature scientific has not provided us the completeand precise causes of this phenomenon. We would like to analyze whetherthere exists long-term causality between temperature and carbon dioxidewith the Vostok Ice-Core data from 400,000 years ago to 6,000 years ago.According to the properties of data, intervention analysis and time seriesanalysis with exponential time trend are applied as the methods of univariateanalysis, so that we can fit the deterministic time trends of data,and we select the time series model with exponential time trend as the finalmodel of univariate analysis by comparison. With the idea of Perron in1989, we consider determining the structural break points based on the natureof data, in order to find the best model. Then using a bivariate systemto analyze both temperature and carbon dioxide simultaneously will bereasonable, and the Granger-Causality test is an appropriate one to judgethe forecasting relation between these two series. Finally, the null hypothesisof the test, that carbon dioxide does not Granger-cause temperature,should be rejected, and this is a one way relation. Thus, this article canprovide strong empirical evidence to show that carbon dioxide causes thechange of temperature.

No description available.

In this thesis, we consider measures of association for ordinal variables from a theoretical perspective. In particular, we study the phi-coefficient, the tetrachoric correlation coefficient and the polychoric correlation coefficient. We also introduce a new measure of association for ordinal variables, the empirical polychoric correlation coefficient, which has better theoretical properties than the polychoric correlation coefficient, including greatly enhanced robustness. In the first article, entitled ``On the relation between the phi-coefficient and the tetrachoric correlation coefficient'', we show that under given marginal probabilities there exists a continuous bijection between the two measures of association. Furthermore, we show that the bijection has a fixed point at zero for all marginal probabilities. Consequently, the choice of which of these measures of association to use is for all practical purposes a matter of preference only. In the second article, entitled ``A generalized definition of the tetrachoric correlation coefficient'', we generalize the tetrachoric correlation coefficient so that a large class of parametric families of bivariate distributions can be assumed as underlying distributions. We also provide a necessary and sufficient condition for the generalized tetrachoric correlation coefficient to be well defined for a given parametric family of bivariate distributions. With examples, we illustrate the effects on the polychoric correlation coefficient of different distributional assumptions. In the third article, entitled ``A generalized definition of the polychoric correlation coefficient'', we generalize the polychoric correlation coefficient to a large class of parametric families of bivariate distributions, and show that the generalized and the conventional polychoric correlation coefficients agree on the family of bivariate normal distributions. With examples, we illustrate the effects of different distributional assumptions on the polychoric correlation coefficient. In combination with goodness-of-fit p-values, the association analysis can be enriched with a consideration of possible tail dependence. In the fourth article, we propose a new measure of association for ordinal variables, named the empirical polychoric correlation coefficient. The empirical polychoric correlation coefficient relaxes the fundamental assumption of the polychoric correlation coefficient so that an underlying joint distribution is only assumed to exist, not to be of a particular parametric family. We also provide an asymptotical result, by which the empirical polychoric correlation coefficient converges almost surely to the true polychoric correlation under very general conditions. Thus, the proposed empirical polychoric correlation coefficient has better theoretical properties than the polychoric correlation coefficient.

No description available.

We have by Statistics Sweden (SCB) been given the task of using different dynamic regression models in order to forecast service price index for sea transport. The aim is to see whether these models provide better forecasts than those previously used. This essay aim to identify, estimate and evaluate the selected prediction models.   Through our data material we were given access to 28 sightings of sea transport index during the period of 2004 q1 to 2010 q4. We have chosen to evaluate three different transfer function models, one ARIMA model and one naive forecasting model. The input variables we decided to test in our transfer function models were the price of petroleum products, the port activity in Swedish ports and the lending rate of Swedish Central bank.   The results of our study suggest that transfer function models generally provide better models than the ARIMA model and the naive forecast model. Results also show that both the transfer function models and ARIMA model seem to provide better models than the naïve forecasting model.  The transfer function model that gave the lowest forecasting errors had interest rate as an input variable.

No description available.

No description available.

No description available.

Criminality can have serious and numerous effects of both individuals and society and why youths engage in criminal actions is a question concerning several instances. In this study we investigate possible factors that influence criminality among youths by analyzing a self-survey questionnaire. To identify the factors the technique of exploratory factor analysis is used. The relationships between the factors and criminality are then assessed with a structural equation model. The objective is also to investigate if the factors relate differently to criminality depending on gender why the structural equation model is performed on males and females separately. From the analysis we identified seven factors. Four of them seem to have an impact on youths’ criminal behavior; criminality among friends, attitudes towards severe crimes, relationships to friends and exposure and participation in negative social behavior. For the other three factors (school performance, family relations and attitude towards illicit actions) we found no clear relationship to criminal behavior. The results also indicate that criminality among friends influence males to a larger degree than females. Whereas the relationship between criminal behavior and exposure and participation in negative social behavior is stronger for females than males. For the other factors no major differences between male and females can be distinguished. The overall picture produced by the analysis suggest that the influences of the identified factors on criminality are quite modest and that the patterns for males and females are quite similar.

Diss. (sammanfattning) Linköping : Linköpings universitet, 2007. / Härtill 4 uppsatser.