This document contains the development of a methodology for validation of data obtained through an Inertial Measurement Unit s (IMU) readings when coupled to an ordinary bicycle. The IMU s readings may have high sensibility to the tests regime and/or experimentation conditions (to be better explained later, in detail), especially when it comes to external factors, so making them susceptible to high variance uncertainties; thus resulting in readings (signals) corrupted by impulsive noise. The implemented methodology consists in establishment of a reference model generated by prior knowledge of the experiments conditions for then establishing causality between distinct readings, statistical analysis of estimated waveforms and recovery of desired signals from the corrupted, observed data. In conclusion there proposes, after detailing the mathematical methodology for data validation, the construction of an apparatus for testing in a controlled environment, where the further implementation of the designed software (programmed in MATLAB environment) will be strictly needed for the analysis of newly obtained data.