It is a consensus that stiction brings strong oscillations for chemical plants and its automatic detection still needs efforts to be fulfilled. Despite the large number, no stiction detection method covers with confidence all cases found industrially. Conditions such as strong noise presence, disturbance, or low sampling may lead to misclassification of the oscillation cause. In this work, a technique for dealing with low sampling signals is introduced. The technique is based on grouping different periods of the signal in such a way that the resulting signal exhibits a higher number of data per cycle and can be analyzed by conventional stiction detection methods. The technique is applied to simulated and industrial data. Results demonstrate the best detection performance after the application of the proposed technique, where stiction detection becomes virtually independent of the sampling period.