engleski

Affective reactions: What can be recorded?

Elicitation of startle reflex and affective reactions by visual, auditory and tactile stimuli, as well as synchronized measurement of the corresponding physiological, speech and facial responses has been conducted in healthy participants, using a developed laboratory system. Ten female first-year university students passed through five elicitation paradigms: (1) acoustic startle probes lasting 40 ms, with intensity 108 dB (A) SPL ; (2) airblasts to the occiput lasting 250 ms, using air pressure of 140 p.s.i. ; (3) loud aversive sounds lasting 1 s ; (4) aversive images lasting 2 s ; and (5) simultaneous deliveries of aversive images, sounds and airblasts having the same properties as in the prior respective paradigms. Each paradigm lasted two minutes and included three pseudo-randomly ordered presentations of stimuli and three occasions when stimuli were omitted. In order to capture the impact of stimuli on speech responses, the participants were given on-screen visual reminder to start sustained phonation of vowel ‘a’ shortly before the stimulus could potentially occur. All participants went through the same paradigms in the same order, 1 through 5, with approximately three-minute breaks between successive paradigms. Recorded physiological, speech and facial features have been analyzed for response differences between stimulus and no- stimulus conditions within each paradigm, as well as to reveal potential differences in responding to various paradigms. Among recorded physiological features, skin conductance response showed the largest differences between stimulus and no-stimulus conditions. The most prominent stimulus/no- stimulus differences among speech features were observed in the energy (intensity) of the voice signal and fundamental frequency (F0), while the major stimulus/no-stimulus differences among facial features occurred in eye blinks, eye gaze direction and head distance from camera. Physiological, speech and facial responses to image stimuli (paradigm 4) had the lowest intensity ; this paradigm was the only one that did not elicit startle response at all. The largest median responses in physiological and speech modality were obtained during paradigm 5 (combination of stimuli), and in facial modality during paradigms 2 and 5. Overall, paradigm 5 can be regarded as the paradigm that elicited the most intense responses across all modalities. The tested paradigms generated significant physiological, speech and facial responses to the presented aversive stimuli in the selected sample of participants. The obtained results are encouraging for further work on the laboratory system and elicitation paradigms for multimodal affective responding, in the context of the authors’ research spanning stress resilience and affective computing topics.

Startle; affective reactions; physiological; speech and facial features; airblast; image; sound

nije evidentirano