Re: Sensitivity to increments and decrements

["Alain de Cheveigne'" <Alain.de.Cheveigne@xxxxxx>]

Hi Christian,

Chait et al. recently found an asymmetry in 
reaction times and brain responses for 
transitions between increases and decreases in 
interaural correlation (a rough binaural analogue 
of  monaural intensity change).

Such asymmetry is to be expected if the internal 
representations of the states to be discriminated 
have different variability. If the internal 
representation of A is more variable than that of 
B, an A-->B fluctation is more likely to occur by 
chance than a B-->A fluctuation, leading to a 
higher probability of false positive.  If the 
decision mechanism adapts its criteria to 
equalize false positive rates in both directions, 
a larger step will be required for detection in 
one direction than the other.  You also expect 
the same asymmetry if the stimulus states (rather 
than their representation) differ in variability.

For intensity, you might expect an asymmetry if 
intensity were coded, say, by the spike rate of a 
Poisson process. Spike count variance is 
proportional to spike count mean, leading to the 
prediction that an increment is easier to detect 
than a decrement.  Is that the case?

Alain


Chait, M., Poeppel, D., de Cheveigné, A. and 
Simon, J. Z. (2007). "Processing asymmetry of 
transitions between order and disorder in human 
auditory cortex." J. Neurosci. 27: 5207-5214.

> Dear List,
>
> Is there anything known about the existence of 
> differences in the sensitivity to intensity 
> increments versus to intensity decrements?
>
> Laurent Demany pointed me to a paper by Sinnott 
> et al. (1985) who found no such difference in 
> humans, while they found an advantage for 
> increments in monkeys:
>    Sinnott, J. M., Petersen M. R., Hopp, S. L. (1985).
>    Frequency and intensity discrimination in humans and monkeys.
>    Journal of the Acoustical Society of America, 78, 1977-1985.
> Is this finding (as to the symmetry of human 
> increment / decrement sensitivity) unchallenged?
>
> Best regards,
> Chris
>
> --
> Prof. Dr. Christian Kaernbach
> Allgemeine Psychologie
> Institut für Psychologie
> Christian-Albrechts-Universität zu Kiel
> Olshausenstr. 62
> D-24098 Kiel
> Germany
> www.kaernbach.de