THE PULSE DURATIONS OF GAMMA-RAY BURSTS (GRBS) HOLD THE KEY TO CRUCIAL INFORMATION IN UNDERSTANDING THE UNDERLYING PHYSICS. IN PARTICULAR THE WIDELY-ADOPTED CLASSIFICATION OF LONG AND SHORT GRBS WITH THE SEPARATION OF ~ 2 SEC IN THE BURST DURATIONS HAS BEEN COMMONLY USED TO INFER THE DIFFERENT ORIGINS OF GRBS. HOWEVER THE OBSERVED BURST DURATION CAN SUFFER FROM DIFFERENT DEGREES OF INSTRUMENTAL AND OBSERVATIONAL BIASES AND THUS MIGHT NOT REFLECT THE TRUE INTRINSIC DURATION. WE PROPOSE TO STUDY THE BIASES OF THE BURST DURATIONS IN THE GRB SAMPLE FROM THE SWIFT BURST ALERT TELESCOPE (BAT) AND QUANTIFY THE INSTRUMENTAL EFFECTS AND SYSTEMATIC UNCERTAINTIES ON THOSE DURATIONS. THE PROPOSED STUDY WILL UTILIZE A CODE WE DEVELOPED THAT IS CAPABLE OF SIMULATING SWIFT/BAT GRBS WITH ACCURATE INSTRUMENTAL RESPONSE AND THE BAT TRIGGER ALGORITHM.