I present a simple metric based on a band-pass-filtered estimate of the variance of Kepler lightcurves for red giant stars, which I am using as a way to understand differences observed in the seismic properties of red giant branch and red-clump stars, and also to infer the binary fraction of the Kepler red giant branch stars.

Simulated power spectrum data have been created for synthetic Kepler fields, with a population of stars of different binary fractions generated by the TRILEGAL code, using scaling relations for the asteroseismic parameters, simple approximations for the granulation background, and accounting for mode visibilities and shot noise. Analyses of the artificial data are being used to guide my interpretation of the real data. When applied to Kepler light curves, the variance metric discriminates between red giant branch stars and those in the red clump, however in the simulated fields, this is not present, presumably due to the departure of these more evolved stars from the scaling relations in reality. I shall also discuss how it may be possible to infer the underlying binary fraction of the population.