OBJECTIVE DETECTION OF AUDITORY-EVOKED POTENTIALS - COMPARISON OF SEVERAL STATISTICAL TESTS IN THE FREQUENCY-DOMAIN ON THE BASIS OF NEAR-THRESHOLD ABR DATA
E. Sturzebecher et M. Cebulla, OBJECTIVE DETECTION OF AUDITORY-EVOKED POTENTIALS - COMPARISON OF SEVERAL STATISTICAL TESTS IN THE FREQUENCY-DOMAIN ON THE BASIS OF NEAR-THRESHOLD ABR DATA, Scandinavian audiology, 26(1), 1997, pp. 7-14
A fully objective electric response audiometry (ERA) requires an objec
tive response detection by an appropriate statistical test. The Raylei
gh test, Watson's U-2 test, Kuiper's test and Hodges-Ajne's lest check
the phase angle distribution of a Fourier harmonics. The modified Ray
leigh test uses, in addition to the phase angles, the amplitude inform
ation in the form of the ranks of the spectral amplitudes, whereas mag
nitude-squared coherence (MSC) uses the spectral amplitudes themselves
. The signal detection performance of these six tests was judged on th
e basis of a sample of near-threshold click-evoked ABR. MSC was found
to be the best suited test out of the six tests investigated, but the
performance differences to the modified Rayleigh test (and even to the
unmodified Rayleigh test), and to Watson's test as well were only sli
ght. Hodges-Ajne's test and Kuiper's lest have the lowest sensitivity
and the mean time required for response detection is longest for Hodge
s-Ajne's test.