טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
Ph.D Thesis
Ph.D StudentUrit Gordon
SubjectDynamic Perspective on Contextual Effects of Visual
and Auditory Perception - A Psychophisical Study
DepartmentDepartment of Medicine
Supervisors Full Professor Brenner Naama
Full Professor Marom Shimon
Full Thesis textFull thesis text - English Version


Abstract

This is a psychophysical study which relates to the gap between a physical stimulus and the subjective perceptual experience it evokes. It is concentrating on the issue of fluctuations in perception of weak stimuli in the visual and the auditory modalities.

Fluctuations mean that repeated presentations of the same input level, on the detection threshold, result in different responses. Streams of these responses are characterized by structural features and functional regularities which imply that these fluctuations are not low-level noise but are connected to cognitive processes.

In this study we specifically inspect the dynamic properties of these responses.

The dynamic perspective and, accordingly, experimental design in this work are not

conventional. Unique methodologies of analysis were developed and adapted to match the dynamic properties of interest while the limitations of these methodologies were also considered. We approached the dynamics in two ways a) directly- by manipulating input temporal structure and b) indirectly by inspecting the linkage between perception processes of audition and vision.

In order to investigate the temporal context in which the stimuli are presented, we

have chosen to refer to the changes in the input amplitude around the sensing threshold.

We presented three structures which differ in their level of internal correlations: no

correlation, correlated and highly correlated. These structures, which were chosen

to resemble different natural rhythms, enabled us to evaluate general concepts; We

showed that opposing tendencies take effect over different timescales, with a different balance between the tendencies emerging in each of the regimes, which enabled us to explain the complex structure of any of the response streams. A general manifestation of this balance was formulated into a mathematical model which described well the experimental results in all regimes.

The indirect method revealed some of the complex relations between responses of the two modalities. Although in our experiment auditory and visual stimuli were not bound to a single object, nor they were presented simultaneously, still their responses were linked in their dynamical properties. The specific linkage is dependent on the contextual relations of both modalities with the perceived object. Contextual relations were controlled via a "response clamp" which is a closed loop procedure including ongoing adjustment of the input level reacting to the history of responses. The cross- modal influence was asymmetric, as clamping just the auditory fluctuations was also effected by quenching the visual fluctuations, but this was not the case the other way round.

We have also showed that the properties of perception of a single modality can alter when more modalities are involved. This emphasizes the need to relate to all the contexts in which perception processes are measured, and to the manner in which they are analyzed and interpreted.


In general, the study supports the view that the dynamics of response fluctuations

are related to cognitive processes. We demonstrated the importance of relating to temporal and other contexts in psychophysical measurements and in their interpretations.

The novel dynamic point of view also contributed to developing new methods that allow the relations between stimulus, response and context to be examined over various timescales.