Scientific and Information Visualization have evolved into two sub disciplines of visualization that are quite distinct from each other. A separation between these fields was justified because of significant differences in the types of data sets --- spatial versus and non-spatial --- and corresponding problems tackled by each. This separation led to a gap between the software tools, hardware technologies and visual languages adopted. However, available graphics and visualization techniques, regardless of their intended purpose, can be applied advantageously to both types of data sets. In this context, 3D graphics rendering methods, which are readily adopted for Scientific Visualizations, can be used to enhance 2D Information Visualization techniques. Similarly, certain spatial data sets can be better visualized with 2D Information Visualization approaches to support a range of tasks such as comparison and detection of patterns.

The goal of this thesis is to develop Scientific and Information Visualizations with the following motivations: 1) dimensionality of a representation should be defined and addressed not as a binary distinction between 2D and 3D, but rather as a continuum such that visualizations can have intermediary dimensionality, 2) dimensionality of a visual representation should be determined by the task at hand rather than the spatial properties of the data or the conventions of the respective visualization fields.

This thesis will present evidence from case studies that support the aforementioned propositions. It will discuss two case studies --- Stereoscopic Highlighting and Contour Maps --- that illustrate how 3D elements can be used to enhance 2D visualizations of non-spatial data. The benefits of less dimensional representations will be explored via LineSets and Weighted Graph Comparisons for Brain Connectivity Analysis case studies. Our results show that more compact representations can be realized and key tasks such as visual comparisons can be carried out more effectively with less dimensional visual representations of spatial data.