Significant progress in the description of quasar variability has been recently made by employing SDSS and POSS data. Common to most studies is a fundamental assumption that photometric observations at two epochs for a large number of quasars will re
veal the same statistical properties as well-sampled light curves for individual objects. We critically test this assumption using light curves for a sample of $sim$2,600 spectroscopically confirmed quasars observed about 50 times on average over 8 years by the SDSS stripe 82 survey. We find that the dependence of the mean structure function computed for individual quasars on luminosity, rest-frame wavelength and time is qualitatively and quantitatively similar to the behavior of the structure function derived from two-epoch observations of a much larger sample. We also reproduce the result that the variability properties of radio and X-ray selected subsamples are different. However, the scatter of the variability structure function for fixed values of luminosity, rest-frame wavelength and time is similar to the scatter induced by the variance of these quantities in the analyzed sample. Hence, our results suggest that, although the statistical properties of quasar variability inferred using two-epoch data capture some underlying physics, there is significant additional information that can be extracted from well-sampled light curves for individual objects.
