[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Instruction for subjects



Dear Fatima and List,

One way to deal with the fact that only one of the sounds to be compared
has a noise burst in the middle while the other has not, is to make BOTH
sounds in the comparison have a noise burst.  One sound would be the one
you described -- an FM sweep with a white noise burst (WNB) replacing the
gap.  The one with which it is to be compared could also have a noise burst
in the middle, but one that doesn't produce continuity.  This can be
accomplished if the noise is a notched-noise burst (NNB).  The NNB can be
made as loud as the WNB, but it shouldn't have any frequencies at or near
the frequency of the tone glide at its midpoint.  Of course the effects of
higher and particularly lower frequencies in the NNB  will spread; so you
will have to do some preliminary listening to discover how wide the notch
has to be in NNB in order not to induce continuity. The subjects can then
be told to ignore the noise bursts (which may not sound the same) in both
sounds, and to judge only the tone glides.

You might consider having the following conditions:
1.  Actually-continuous glide passing through WNB.
2.  Glide with gap filled by WNB (auditory induction condition)
3.  Actually-continuous glide passing through NNB
4.  Glide with gap filled by NNB.

The percepts resulting from these conditions should be:
1.  Continuous glide
2.  Continuous glide
3.  Continuous glide
4.  Discontinuous glide (verification of the fact that NNB doesn't produce
continuity)

Any differences between the sounds of 1, 2, and 3 would be very
interesting.

If condition 2 is not actually producing continuity (according to your own
ears), you might try:
-- A shorter gap
-- Having, as the stimulus, a series of connected ascending and descending
glides, with the gaps (or noise) placed in the middle of each ascending and
each descending glide.  This repetition may favor continuity.

Another change you might consider is in the question you ask.  If there are
two sounds to be compared, S1 and S2, you might say: "Ignore the noises and
focus only on the tones.  Which tone has a longer gap  in it, the first or
the second.  If Condition 4, above, really sounds discontinuous, it could
be used as S1, but you could change the duration of the gap on different
trials (keeping the duration of the noise constant).  In this way you could
calibrate the size of any induction effects, even if they only shortened
the apparent length of the gap, rather than eliminating it altogether.

Dick Warren has a lot of experience in these types of experiment and much
of what I have proposed is based on his findings.  You might want to
consult him directly.  I'm sorry, but I don't have his current e-mail
address.  Maybe other list members do.  Dick, are you out there?

Best of luck,

Al

----------------------------------------------------
Albert S. Bregman, FRSC
Emeritus Professor
Psychology Dept., McGill University
1205 Docteur Penfield Avenue
Montreal, Quebec
Canada  H3A 1B1

Voice & Fax: +1 (514) 484-2592
----------------------------------------------------
----- Original Message ----- 
From: "FATIMA HUSAIN" <fthusain@xxxxxxxxxxxxx>
To: <AUDITORY@xxxxxxxxxxxxxxx>
Sent: Tuesday, October 25, 2005 1:42 PM
Subject: Instruction for subjects


> Dear List
>
> I am running some pilot behavioral studies of auditory continuity
illusion
> (aka temporal induction).
> The sounds are FM sweeps with gaps or noise-inserted gaps.
> For various reasons I am running them as "same/different" trials, wth the
> first sound of the trial being a regular FM sweep and the second one
being
> the modified FM sweep.
>
> When I ask the subjects, "Is the second sound as continuous as the first
> sound?" they become confused.
>
> Even FM sweeps with much-louder noise inserted in the gap are considered
> by some subjects to be not very continuous. (contrary to literature)
>
> How do I change my instructions (basically, the confusion is around the
> word 'continuous')? Or change the task?
>
> Any ideas?
>
> Thanks
>
> Fatima T. Husain, Ph.D.
> NIDCD/NIH
> Bldg. 10/ Room 8S-235D
> Phone: 301-594-7758
>
>