The various projects presented in this thesis contribute to the expansion of our knowledge and understanding of the fundamental processes involved in the formation of a galaxy and its stellar populations. For this purpose, two different techniques are applied: A photometric study of the properties of the stellar halo surrounding the edge-on galaxy NGC 3957, and spectroscopic studies of the chemical compositions of stars in the nearby dwarf spheroidal galaxies Sextans, Fornax, and Sculptor. Due to their extreme weakness, studies of halos outside the Local Group are very challenging, and available data are still very scarce with limited quality. The results of previous studies are quite puzzling. Lequeux et al. (1996, 1998); Zibetti et al. (2004); Zibetti & Ferguson (2004), for example, report the discovery of external stellar halos with extremely red integrated colors, which were neither in agreement with the halo stars observed in the Milky Way or M31, nor with any other realistic stellar population. To investigate this problem, we obtained deep images of the edge-on galaxy NGC 3957, to extract minor axis surface brightness profiles, allowing the study of the halo properties in regions which are not contaminated by the much brighter disk or bulge component. While the central regions of our profiles are very well described by a de Vaucouleurs law, we find a clear excess of light at larger radii, indicating the presence of a stellar halo. Our color profiles rise to very red colors at large radii, however, a careful error analysis shows that systematic errors are mainly responsible for this rise. In the radial range where the halo component is dominant and the systematics are still small, our results are compatible with the expected integrated halo colors of the Milky Way, M31 and the slightly more distant galaxy NGC 891. Even though this is no proof, it suggests that the stellar population of the halo of NGC 3957 is a rather typical one. It is shown that the limiting factor is definitely a sufficiently accurate determination of the background level. We propose a few ways how this could be improved. Nevertheless, the sky subtraction remains the largest challenge for future studies of this type, especially if the goal is to trace the stellar halo to larger radii, comparable to the halo of the Milky Way or M31. The targets for the spectroscopic analysis of stars in dwarf galaxies were selected from an extended low resolution survey of the Ca II triplet (CaT) region, containing ∼2000 stars in Sextans, Sculptor, and Fornax. This is the largest survey of this kind done so far targeting stars outside the Milky Way. One of the most striking results of this survey was the absence of stars with metallicity estimations from the strength of the CaT lines which were below -2.8 (Helmi, 2006). This is particularly interesting, since the Milky Way halo contains a wealth of stars clearly below this limit (Cayrel et al., 2004). Such a difference would have consequen