Abstract: Using several time-series of observed neutron scattering data from a nematic liquid crystal, we analyze their frequency spectra and subsequently reconstruct the observed time series. Several approaches are tried: the conventional power-spectrum analysis; the maximum-entropy method; and Bayesian spectrum analysis. We also study the influence of random noise, and test the reliability of the Hurst-exponent analysis.