To United Minds, large-scale surveys are an important offering to clients, not least the public opinion poll Va?ljarbarometern. A risk associated with surveys is satisficing ? sub-optimal response behaviour impairing the possibility of correctly describing the sampled population through its results. The purpose of this study is to ? through the use of multivariate outlier detection methods - identify those observations assumed to be non-representative of the population.

The New National Elevation Model (NNH) is a new high-resolution digital elevation model (DEM) of Sweden from airborne laser scanning. It creates many new opportunities, particularly in the area of flood mapping. NNH is provided by Lantmäteriet in two formats, both in raw LIDAR (Light Detection and Ranging) data and in grid format with two meter resolution. These alternatives have advantages and disadvantages and the aim of this thesis research is to identify these. One of the focuses of the study is data storage and thus data structure analysis, resolution and storage facilities. The research questions are:Why and in what context the different NNH-products from the National Land Survey are used (DEM 2+ or point cloud)?What constraints and opportunities are created by the different options, mainly in terms of different software, resolution, and data storage?The study area is Kärsön in Ekerö municipality located in Stockholms län and has an approximate area of 25 square kilometers. The study is divided into two parts. The first objective is to identify the consequences of using different software to create DEM from pointcloud compared to the DEM2+ model. Height models with a two meter resolution are created in FME and ArcGIS. The models are then compared with the grid from Lantmäteriet, created in TerraScan. The second objective is to examine the impact of the change in resolution, both the storage aspect and both the accuracy aspect. Inverse Distance Weighted (IDW) is an interpolation method which in previous studies proved to have the best results on high resolution LIDAR data. This model was tested and compared with a model from FMEs built-in function and the model from Lantmäteriet wich are based on triangulation (also proved a good method in previous studies). The grid created in TerraScan has good properties such as accuracy.

Common wisdom dictates that because payment of dividends decreases funds available for investment, a high payout ratio is to be associated with low earnings growth. This notion has been challenged in recent years by empirical research that has found a positive relationship between payout ratio and future earnings growth, both on the aggregated market level and the company level in different countries. This study investigated the aforementioned relationship on the Swedish market by statistically analyzing future earnings growth as a function of payout ratio on a large sample of publicly listed companies over the period 1980 to 2011. The results are in line with those found in previous research and indicate that companies with high dividend payout ratio tend to experience strong future earnings growth. The results are robust to the presence of mean reversion, the use of an alternative accounting measure of earnings, non-symmetrical earnings growth cycles and alternative practices for dealing with Outliers..

Denna experimentella studie syftar till att undersöka huruvida en eventuell träningseffekt i frekvensdiskrimination kvarstår efter en tids träningsuppehåll. Tidigare forskning har visat tydliga samband mellan frekvensdiskriminering och talutveckling, kortikal signalbehandling och dyslexi. Det har klargjorts genom tidigare studier att förmågan att diskriminera frekvenser är möjligt att träna upp. Därmed är träning i frekvensdiskrimination en lämplig del i en behandlingsmetod med syfte att utveckla den auditiva perceptionen. Denna studie undersöker de praktiska förutsättningarna till denna behandlingsmetod genom att utreda varaktigheten av den eventuella träningseffekten 14 dagar efter träningen.