The three-point correlation function (3PCF) for galaxies provides an opportunity to measure the non-Gaussianity generated from nonlinear structure formation and also probes information about galaxy formation and evolution that is generally not available from the two-point correlation function (2PCF). We measure the 3PCF of the Sloan Digital Sky Survey DR7 main sample galaxies in both redshift and projected spaces on scales up to 40Mpc/h. We explore the dependence of the 3PCF on galaxy stellar mass and color in order to constrain the formation and evolution for galaxies of different properties. The study of the dependence on these properties also helps better constrain the relation between galaxy stellar mass and color and the properties of their hosting dark-matter halos. We focus on the study of the reduced 3PCF, Q, defined as the ratio between the 3PCF and the sum of the products of the 2PCFs. We find a very weak stellar mass dependence of Q in both redshift and projected spaces. On small scales, more massive galaxies tend to have slightly higher amplitudes of Q. The shape dependence of Q is also weak on these small scales, regardless of stellar mass and color. The reduced 3PCF has a strong color dependence for the low-mass galaxies, while no significant dependence on color is found for the high-mass galaxies. Low-mass red galaxies have higher amplitudes and stronger shape dependence of the reduced 3PCF than the blue galaxies, implying that these low-mass red galaxies tend to populate filamentary structures. The linear galaxy bias model fails to interpret the color dependence of Q, emphasizing the importance of a nonvanishing quadratic bias parameter in the correct modeling of the galaxy color distribution.

We have accurately measured the bispectrum for four scale-free models of structure formation with the spectral index $n=1$, 0, -1, and -2. The measurement is based on a new method that can effectively eliminate the alias and numerical artifacts, and reliably extend the analysis into the strongly non-linear regime. The work makes use of a set of state-of-the art N-body simulations that have significantly increased the resolution range compared with the previous studies on the subject. With these measured results, we demonstrated that the measured bispectrum depends on the shape and size of $k$-triangle even in the strongly nonlinear regime. It increases with wavenumber and decreases with the spectral index. These results are in contrast with the hypothesis that the reduced bispectrum is a constant in the strongly non-linear regime. We also show that the fitting formula of Scoccimarro & Frieman (1999) does not describe our simulation results well (with a typical error about 40 percent). In the end, we present a new fitting formula for the reduced bispectrum that is valid for $-2 \leq n \leq 0$ with a typical error of 10 percent only.

We present velocity statistics of galaxies and their biases inferred from the statistics of the underlying dark matter using a cosmological hydrodynamic simulation of galaxy formation in low-density and spatially flat cold dark matter cosmogony. We find that the pairwise velocity dispersion (PVD) of all galaxies is significantly lower than that of the dark matter particles, and that the PVD of the young galaxies is lower than that of the old types, and even of all galaxies together, especially at small separations. These results are in reasonable agreement with the recent measurements of PVDs in the Las Campanas redshift survey, the PSCz catalogue and the SDSS data. We also find that the low PVD of young galaxies is due to the effects of dynamical friction as well as the different spatial distribution. We also consider the mean infall velocity and the POTENT density reconstruction that are often used to measure the cosmological parameters, and investigate the effects of spatial bias and dynamical friction. In our simulation, the mean infall velocity of young galaxies is significantly lower than that of all the galaxies or of the old galaxies, and the dynamical bias becomes important on scales less than 3Mpc/h. The mass density field reconstructed from the velocity field of young galaxies using the POTENT-style method suffers in accuracy both from the spatial bias and the dynamical friction on the smoothing scale of R_s=8Mpc/h. On the other hand, in the case of R_s=12Mpc/h, which is typically adopted in the actual POTENT analysis, the density reconstruction based on various tracers of galaxies is reasonably accurate.

For the first time we calculate quantitatively the influence of inhomogeneities on the global expansion factor by averaging the Friedmann equation. In the framework of the relativistic second-order Zel'dovich-approximation scheme for irrotational dust we use observational results in form of the normalisation constant fixed by the COBE results and we check different power spectra, namely for adiabatic CDM, isocurvature CDM, HDM, WDM, Strings and Textures. We find that the influence of the inhomogeneities on the global expansion factor is very small. So the error in determining the age of the universe using the Hubble constant in the usual way is negligible. This does not imply that the effect is negligible for local astronomical measurements of the Hubble constant. Locally the determination of the redshift-distance relation can be strongly influenced by the peculiar velocity fields due to inhomogeneities. Our calculation does not consider such effects, but is contrained to comparing globally homogeneous and averaged inhomogeneous matter distributions. In addition we relate our work to previous treatments.

We show that in a hierarchical clustering model the low-order statistics of the density and the peculiar velocity fields can all be modelled semianalytically for a given cosmology and an initial density perturbation power spectrum $P(k)$. We present such models for the two-point correlation function $\xi(r)$, the amplitude $Q$ of the three-point correlation function, the mean pairwise peculiar velocity $< v_{12}(r)> $, the pairwise peculiar velocity dispersion $< v_{12}^2(r)>$, and the one-point peculiar velocity dispersion $< v_1^2 >$. We test our models against results derived from N-body simulations. These models allow us to understand in detail how these statistics depend on $P(k)$ and cosmological parameters. They can also help to interpret, and maybe correct for, sampling effects when these statistics are estimated from observations. The dependence of the small-scale pairwise peculiar velocity dispersion on rich clusters in the sample, for instance, can be studied quantitatively. There are also significant implications for the reconstruction of the cosmic density field from measurements in redshift space.