Introduction

Quasar evolution and selection bias

Radio-quiet quasars have historically been selected by an ultraviolet excess (UVX) method, e.g., the Palomar-Green Bright Quasar Survey. Such surveys have been fundamental in determining the selection criteria for high-redshift quasars (e.g., from the Sloan Digital Sky Survey), yet the role in which dust extinction can bias such surveys is still unclear, and therefore the observed number density of distant quasars selected by such methods should be treated with caution. Indeed, radio-loud quasars (which are selected independent of such biases) have a larger spread of optical and infrared spectral indices than their optically-selected counterparts (see Figure; from Simpson & Rawlings 2000). Warren, Hewett, &smp; Foltz (2000) are attempting to address this with a new quasar-selection technique which they call the KX method. Our imaging data will allow us to construct samples of optical/infrared-selected quasars candidates for spectroscopic follow-up (we predict ~40 quasars with z>2.2 and K<19), which will be compared with the quasars discovered from the X-ray identifications. While the catalogs will have a substantial overlap, the quasars which appear in only one catalog will reveal the true spread in optical-infrared-X-ray properties of quasars. This is essential in determining the evolution of the quasar number density, which in turn has implications for the reionization of the Universe and the formation of massive galaxies.