We have analyzed the magnetic susceptibility () measurements for more than 5000 samples of drill cuttings from oil/gas wells in the Venezuelan oil fields to facilitate information extraction and improve data presentation, including the determination of threshold values used in hydrocarbon exploration. The -values for each well typically ranged over several orders of magnitude. We have developed a new scale, called the magnetization contrast magnitude scale , based on a logarithmic transformation of , to better represent the data. Collectively, the -data appear to be influenced by more than one process/source, resulting in skewed bimodal distributions with a significant degree of overlap between the individual modes. Each mode or subpopulation of the -data could be modeled with a lognormal distribution. We have determined that a method based on normal probability plots could be used to decompose the observed bimodal distributions into their individual components and to obtain initial estimates for the statistical parameters that characterize each of these components (i.e., mean , standard deviation , and percentage). Finally, more accurate component parameter estimates were obtained by applying an expectation-maximization fitting algorithm to find the maximum likelihood estimates of the parameters in a proposed Gaussian mixture model. The -value of each sample was partitioned by determining the Gaussian component with the highest posterior probability for each value. Subsequent color display as a drillhole trace of the proportions of the subpopulations, which was interpreted for each sample, could then be used in combination with other logged parameters to reveal information on the spatial distribution of magnetic mineral assemblages that were interpreted to be hydrocarbon related.