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

Re: Online listening tests and psychoacoutics experiments with large N


I am not in online testing, but I find the topic rather interesting.

A few issues pop out:

[1] Variability of presentation conditions. Improvement: equalize signals or
audio playback using data from an initial quick-and-dirty measurement of
detection thresholds at different frequencies. A way can also be found to
roughly tell apart the participant-specific sensitivity from the frequency
response of the audio playback chain.

[2] Attention level. Improvement: interleave experimental trials with
auditory detection trials. If the reaction time in any of the detection
trials exceeds a given acceptable threshold, the experiment is terminated
and data is not considered.

[3] The same person could participate to the experiment more than once. This
might not be ideal. Improvement: [a] the participant gives an email address,
and the system checks whether substrings of the address have already been
submitted for participation; [b] a different http address for the experiment
is sent to the mailbox of each of the participants (e.g., the online
executable is given a different name for each of the participants).



Bruno L. Giordano, Ph.D.
Music Perception and Cognition Laboratory
CIRMMT http://www.cirmmt.mcgill.ca/
Schulich School of Music, McGill University
555 Sherbrooke Street West
Montréal, QC H3A 1E3
Office: +1 514 398 4535 ext. 00900