A permutation fuzzy entropy algorithm is proposed that uses sorting and symbolic methods to improve anti-noise performance of electroencephalogram signals known to be highly sensitive to noise disturbances during collection. It was employed to analyse abnormal event-related potentials of schizophrenics focused on P50 potentials of sensory gating, which is the most common paradigm currently used for analysing schizophrenia. The approach for analysing P50 sensory gating in schizophrenics is presented from twenty-seven schizophrenia patients and twenty healthy controls. The values calculated for the patients under the conditioning and testing stimuli were used to calculate the entropy complexity. Results demonstrate that the approach can be effectively used to analyse sensory gating deficits in patients with schizophrenia and that the algorithm can be satisfactorily be used for analysing electroencephalogram signals.