Dear Roberto, 

Yes its possible. We did some experiments on stress level measurement using electrodermal  activity. The way or method of signal analysis plays crucial role on identifying the human physiological changes. Especially, these kind of signals have linear trend and has to focus our analysis on very low frequency range. I attached some of my paper related to you question. Hope it will help you.

Good luck

We were curious about the same thing and I did a lit review a few years back. FWIW, we decided not to use it. (Note: we were interested in infants/toddlers, and the lit is pretty small compared to adults.) E.g., Harrison et al (2006) found very high variability in neonates in hospital several contexts (doi:10.1016/j.earlhumdev.2005.12.008), and even a heel-prick procedure did not elicit SC increases in more than a few infants! (There was a significant *increase* - not decrease - but driven by a minority of infants.) On the other hand, El-Sheikh 2007 found inter-limb and between-time stability in individual children, which seems promising. Also, another study (0162-3257/06/0500-0517/0) found significant decreases only in a group of children w/ autism in response to seeing a picture of a face w/ direct gaze. Other studies had shown related results...however, the direction of effect is not always consistent with what one would expect based on the physiology! Also, informally, I have a colleague who tried using it and had trouble getting reliable results. (Maybe just an inexperience issue?) Some labs get useable SCR data, but it seems there are multiple possible sources of measurement error, and you have to work hard to get clean data. One final concern is that it's not clear how to map changes in SCR to degree of sympathetic tone change. Perhaps the most you can ever say is whether or not you detected any significant change in sympathetic tone - period. That might be enough for you, but we always have to bear in mind the limitations of any given measure. Good luck!

Thank you very much Dr. Murugappan and Dr. O Deak for their valuable answers. Nowadays we are worried in our work team about the processing of the signal and the target population, thus this comentaries are very helpful to the experimental design. Again thank you for your time. Good day!  

I want to echo the skepticism of the above respondents. Electrodermal response has been used for decades in biofeedback, and if the electrodermal response could yield reliable indicators of either attention or emotional state then that would be have been documented by now. The interest in doing has been there, but the results have always been discouraging. The absence of a supportive literature is not for the lack of trying. The problem is that the electrodermal response exhibits a contingency on the personal history of the testee, on the immediate past, as well as on the current physiological state.

Thank you for your help Dr. Silletti and Dr. Othmer. We are very worried about the possibility of the unpredictable nature of the electrodermal activity, related with the physiological and psychological state of the subject before the test. We think that a minucious ellection of the subjects could avoid the uncertain nature of the physiological value, but the determination of this uncertainity is an important objective of the project. We hope that a minucious filtering and analysis of data could give us a little light about the use of the electrodermal activity as emotional indicator. Good day!

Dear Siegfried, Roberto,

I agree with your last sentence, but would like to add that to me this seems to be a characteristic of attentional and emotional responses as such. Any measure of these constructs will thus show these trait and state-dependent characteristics.

Please see the book of Boucsein "Electrodermal Activity" for a nice overview - also showing the tremendous complexity of EDA research and findings. In my experience EDA is a useful indicator of arousal, though the specificity to emotional arousal compared to physiological arousal is IMO not clear - that is EDA will give you an indication of the activation of the sympathetic nervous system independent of the cause. To get a better picture, e.g. tying response to cause, you need to incorporate knowledge about the context. In psychophysiology experiment design restricts the possible causes. In applied research you might want to use other sensor modalities to check for events that caused EDA changes. Finally, Gedeon expresses my feeling towards physiological sensors in general - you have to be familiar with the signal, on a theoretical and practical level - SNR and confounding variables are limiting the inferences that you can draw in a specific setting.  Good luck!

I would like to echo the above skepticism and the recommendation of Boucsein (that is the only comprehensive work on SCR that I'm aware of). I would add only two things.

1. In my experience, skin conductance has a robust and repeatable component in the startle response (along with EMG/HR etc.) Startle models seem to be more practicable than baseline spontaneous fluctuations. Certainly they are easier to model. If you are providing 'acoustic signals' to participants, you should consider this.

2. What is generally regarded as one of the primary problems with SCR is that it isn't a 'neat' measure of sympathetic outflow, but I was very intrigued by this: 

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3898862/

Dominic Bach has done a great deal of work on SCR in the last few years. He has a MatLab toolbox that is will give you access to the analysis cited in the papers above. It's here:

http://scralyze.sourceforge.net/

Best of luck.

JH

Than you very much, Dr Mühl and Dr. Heathers. Of course I am reviewing the work of Bucsein and clarify some important aspects of this research. We hope that the design of the experiment could stablish rigorously the psicological context and clarify the relationship between the EDA and the relationship between atention and memory. The startle response is a very interesting alternative if we are not able to find reproducible results on EDA.

Roberto, ditto on the above, but be sure to look for Boucsein's 2nd edition (2012). Electrodermal Activity, Springer; New York.

The research on emotion has made some significant advances in the last 8-10 years and I would strongly recommend reading one or more of the following:

Barrett, L. F. (2006a).  Are emotions natural kinds?  Perspectives on Psychological Science, 1(1), 28–58.  doi: 10.1111/j.1745-6916.2006.00003.x

Barrett, L.F. (2006b).  Valence as a basic building block of emotional life.  Journal of Research in Personality, 40(1), 35–55.  doi.10.1016/j.jrp.2005.08.006

Barrett, L. F., & Bliss-Moreau, E. (2009).  Affect as a psychological primitive.  In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 41, pp. 167–218).  New York, NY: Academic Press.  doi:10.1016/S0065-2601(08)00404-8

Additionally, if your thinking about SCR, this is related to the Orienting Response, which has been examined extensively, some using variances in tones, by Robert Barry out of Australia.  See for example:

Barry, R.J. (1996).  Preliminary process theory: towards an integrated account of the psychophysiology of cognitive processes.  Acta Neurobiologiae Experimentalis, 56, 469–484.  Retrieved from http://www.ane.pl/pdf/56055.pdf

Barry, R.J. (2006).  Promise versus reality in relation to the unitary orienting reflex: a case study examining the role of theory in psychophysiology.  International Journal of Psychophysiology, 62(3), 353–366.  doi:10.1016/j.ijpsycho.2006.01.004.

Barry, R.J. (2009).  Habituation of the orienting reflex and the development of preliminary process theory.  Neurobiology of Learning and Memory, 92(2), 235–242.  doi:10.1016/j.nlm.2008.07.007.

Barry, R.J. (2012).  Cortical measures in preliminary process theory of the orienting reflex.  International Journal of Psychophysiology, 85(3), 315.  doi: 10.1016/j.ijpsycho.2012.06.073.

Barry, R. J., & James, A. L. (1981).  Fractionation of phasic responses in a dishabituation paradigm.  Physiology & Behavior, 26(1), 69–75.  doi.10.1016/0031-9384(81)90080-9

I (we) just had a paper published and Barry's work was an eye opening break through in completing that project.  His work would help you understand "WHY" a response is elicited.

Thank you for your time, Dr. Palmatier. The bibliography about emotions are a very valuable start point for this research and for other partners in my work area. I am not familiarized with orienting response, but I think that will be an important related phenomena in the experimental design.

We've done a lot of work with GSR/EDA, and continue to use it for a variety of physiological monitoring applications. I would second James' observation that with a startle paradigm (or any with spaced, discrete stimuli), you would likely get some actionable data. However, in extending and generalizing your work down the road in richer stimulus environments, the signal qualities (slow to respond, slow to return to baseline) of EDA will give you a headache, as will interpretation given the mixture of central, peripheral and endocrine processes that contribute to EDA. Facial EMG might provide a better alternative, with a much faster sampling rate that will give you more flexibility in your research down the line.

Roberto and all, I would also suggest to look at two my papers dealing with mathematical modeling of SCR. You will find a new metrics of SCR, which could be interesting for your work. Good day!

Chapter Assessment of Brain Monoaminergic Signaling Through Mathemat...

Article Interlinked positive and negative feedback loops design emot...

Thank you Dr. Poore for your time. We are considering a lot this paradigm about the GSR paradigm with EDA. Your observations about the signal qualities are absolutely useful for the design of the paradigm. We will consider your suggestion about EMG, probably near the eyes, because in past studies probe to be very useful. 

Thank you Dr. Brankovic for your time. Your article had remarkable useful ideas. The conception of the derivative of EDA signal as a methodology to find hidden information and the use of MATLAB System Identification Toolbox is a great idea for expanding our tools for processing data.

Dear Roberto,

we have done a series of laboratory experiments with traffic noise using the SCL as one of the physilogical measures, and we find SCl the mosr reliable parameter which is highly correlated with the subjective evaluation of the stimuli - in terms of disturbance / dislike, but also in terms of approval.

Please be aware that age and gender are major factors with regard to the baseline as well as to the strength of response of SCL. Further we have found several psychological aspects influencing the strength of SCL response, mainly noise sensitivity and sensation seeking.

You might have a look at two publications  - especially the first one which can be seen in full text at the Noise & Health website.

Good luck!

Article Patterns of physiological and affective responses to vehicle...

Article Psycho-physiological responses to the perception of vehicle ...

Thank you Dr. Notbohm, we will aware about the homogeinicity of the population and the baseline for the basic considerations of EDA. Thank you for your publications, they are very valuable for our work.