The invention relates to an automatic test device and method for an
auditory brainstem response (ABR). On the basis of the calculation of an adaptive averaging method, the average number of records isiteratively increased for the ABR records which are respectively collected under multiple sound intensities, and the
signal to
noise ratio is increased until ABR
signal detection conditions are satisfied, wherein an average curve is obtained according to the ABR records collected in groups and batches, the
time lag of the maximum values of cross-correlation functions between the groups is calculated, and it is judged whether or not there are ABR signals having time locking characteristics according to the fact whether or not the deviation of the
time lag is within a specified range. The iteration is terminated when the ABR signals are detected or no ABR signals are detected after a maximum iteration number is reached. The minimum
sound intensity required to detect the signals is adopted as a hearing threshold value, or the iteration number used under each
sound intensity is subjected to function fitting, and the accurate
sound intensity corresponding to the hearing threshold value isobtained by means of an interpolation method. According to the automatic test device and method, the efficiency of detecting the threshold value is high, the accuracy rate is approximate to that of manual judgment and is more objective, the
repeatability is better, and the repeated collection number of the ABR records can be effectively reduced.