322
W.E Dolphin
one such method is presented in Figure 7.15. This method allows a decision
as to the presence or absence of a response to be based on the statistical
probability that the response at the stimulus modulation frequency is
significantly different from noise at adjacent frequencies (e.g., Wei 1990).
In this approach, the averaged response is Fourier transformed and the
Fourier component corresponding to the stimulus modulation frequency as
well as a number of surrounding frequencies (in the present example the
50 surrounding components, 25 on either side) have been plotted in a polar
form thereby providing both amplitude and phase informatibn. The circle
around the origin is the 95% confidence limit of the noise as evaluated by
the 50 frequency bins around the response. The signal may be considered
present if it falls outside of this confidence limit, that is, there is less than a
5% probability that the obtained response is due to chance. In the case in
which multiple amplitude-modulated frequencies are presented simultaneously, each primary frequency is modulated at a different rate, and, therefore, the response is evaluated independently at each frequency.
The audiogram obtained from a P. crassidens using the multiple component EFR technique has been superimposed on a behaviorally obtained
audiogram in Figure 7.16. The AEP audiogram was obtained using three
1,000,000
1,000
10,000
100,000
Frequency (Hz)
160 .,----,---,----;-,--;-,-,c-,-,-------"-----;,-",,, ,--;-,"-,,"-----'--,---'---," "' ""I iT,- - - , - - - - - , - - , - , - , , . . , . , ,
,..:..:..:-'-----,.
i i i ,
I
I , I I I I !
I
I 111111
.........
&. 140
::J
.....
Q) 120
....
-J
a.. 100
en
CO 80
-0
.........
~ 60
rJ)
c
2 40
c
FIGURE 7.16. Comparison of audiograms obtained using a behavioral approach
with that obtained using the electrophysiological (envelope following response, or
EFR) approach. The EFR was obtained using three separate sets of multitone pair
stimuli (indicated as Series I, II, and III). Series I was comprised of carrier frequencies at 0.2, 1.6, and 8kHz; Series II was comprised of carrier frequencies at
0.4,4.0, and 16kHz; Series III was comprised of carrier frequencies at 0.8,2.0, and
32 kHz. The EFR threshold was estimated as the stimulus intensity at which these
was a statistically significant response at the 90% confidence level.
W.E Dolphin
one such method is presented in Figure 7.15. This method allows a decision
as to the presence or absence of a response to be based on the statistical
probability that the response at the stimulus modulation frequency is
significantly different from noise at adjacent frequencies (e.g., Wei 1990).
In this approach, the averaged response is Fourier transformed and the
Fourier component corresponding to the stimulus modulation frequency as
well as a number of surrounding frequencies (in the present example the
50 surrounding components, 25 on either side) have been plotted in a polar
form thereby providing both amplitude and phase informatibn. The circle
around the origin is the 95% confidence limit of the noise as evaluated by
the 50 frequency bins around the response. The signal may be considered
present if it falls outside of this confidence limit, that is, there is less than a
5% probability that the obtained response is due to chance. In the case in
which multiple amplitude-modulated frequencies are presented simultaneously, each primary frequency is modulated at a different rate, and, therefore, the response is evaluated independently at each frequency.
The audiogram obtained from a P. crassidens using the multiple component EFR technique has been superimposed on a behaviorally obtained
audiogram in Figure 7.16. The AEP audiogram was obtained using three
1,000,000
1,000
10,000
100,000
Frequency (Hz)
160 .,----,---,----;-,--;-,-,c-,-,-------"-----;,-",,, ,--;-,"-,,"-----'--,---'---," "' ""I iT,- - - , - - - - - , - - , - , - , , . . , . , ,
,..:..:..:-'-----,.
i i i ,
I
I , I I I I !
I
I 111111
.........
&. 140
::J
.....
Q) 120
....
-J
a.. 100
en
CO 80
-0
.........
~ 60
rJ)
c
2 40
c
FIGURE 7.16. Comparison of audiograms obtained using a behavioral approach
with that obtained using the electrophysiological (envelope following response, or
EFR) approach. The EFR was obtained using three separate sets of multitone pair
stimuli (indicated as Series I, II, and III). Series I was comprised of carrier frequencies at 0.2, 1.6, and 8kHz; Series II was comprised of carrier frequencies at
0.4,4.0, and 16kHz; Series III was comprised of carrier frequencies at 0.8,2.0, and
32 kHz. The EFR threshold was estimated as the stimulus intensity at which these
was a statistically significant response at the 90% confidence level.