We are interested in how populations of neurons coordinate to encode information from multiple simultaneously presented stimuli. We conjecture that information about each presented stimulus may be aggregated within distinct subpopulations of neurons. We aim to provide some insights to this conjecture by principle component anlaysis (PCA) and factor model. From PCA, we ﬁnd some evidence suggesting distinct subpopulations of neurons may exist depending on the preference of neurons to constituents of simultaneously presented stimuli. This evidence motivates a two-factor random factor Model selected by BIC, which provides a decent estimation for sample covariance matrix. And from the factor loadings matrix, we ﬁnd the latent factors are interpretable, which may relate to the preference of a single neuron on constituents of the simultaneously presented stimuli.