What should I know to use EDA data in my experiment?


What is Electrodermal Activity (EDA)?
Electrodermal activity refers to electrical changes, measured at the surface of the skin, that arise when the skin receives innervating signals from the brain. For most people, if you experience emotional activation, increased cognitive workload or physical exertion, your brain sends signals to the skin to increase the level of sweating. You may not feel any sweat on the surface of the skin, but the electrical conductance increases in a measurably significant way as the pores begin to fill below the surface.
How is EDA measured?
EDA can be measured in many different ways electrically including skin potential, resistance, conductance, admittance, and impedance (see the book Electrodermal Activity by Wolfram Boucsein). The Empatica E4 provides a way to capture electrical conductance (inverse of resistance) across the skin. It achieves this by passing a minuscule amount of current between two electrodes in contact with the skin. The units of measurement for conductance are microSiemens (μS).
What are tonic and phasic changes in skin conductance?
Skin conductance measurement is traditionally characterized into two types – tonic skin conductance level and phasic skin conductance response – which can roughly be thought of as “the smooth underlying slowly-changing levels” vs. “the rapidly changing peaks.”

Figure 1, the top graph depicts a high quality EDA signal. Exemplary skin conductance responses (SCRs) are circled.  These SCR's are examples of "phasic activations".  The tonic value is the more smoothly-changing level, approximated here by the straight white line.

Figure 2 The top graph depicts an EDA signal which is slowly climbing and has no significant phasic activity.

On EDA response to stimuli and SCRs
Phasic – Phasic skin conductance measurements are typically associated with short-term events and occur in the presence of discrete environmental stimuli - sight, sound, smell, cognitive processes that precede an event such as anticipation, decision making, etc. Phasic changes usually show up as abrupt increases in the skin conductance, or “peaks” in the skin conductance. These peaks are generally referred to as Skin Conductance Responses (SCRs).
What is a Skin Conductance Response (SCR)?
Skin Conductance Responses (SCRs) are abrupt increases in the conductance of the skin. The big SCRs that are circled in Figure 1 illustrate their most common form: they usually have a faster rise time than decay time.
On Skin Conductance Level and Baseline:
What is Skin Conductance Level (SCL)?
The SCL usually refers to the raw level of conductance of the skin.  It may also be computed as an average over longer intervals, typically ranging from tens of seconds to tens of minutes. 
Tonic – Tonic skin conductance is generally considered to be the level of skin conductance in the absence of any particular discrete environmental event or external stimuli. The tonic skin conductance level can slowly vary over time in an individual depending upon his or her psychological state, hydration, skin dryness, and autonomic regulation. The baseline tonic skin conductance level may vary from day to day.  Tonic changes in the skin conductance level typically occur in a period of from tens of seconds to minutes.


Control the quality of the EDA signal

Dynamic changes in the electrical properties of the skin have been associated with emotional response since the late 1800's. EDA has been measured in psychophysiology studies in laboratory settings for over 100 years. The Society for Psychophysiological Research has developed recommendations for the optimal recording conditions that are based on the maximum signal-to-noise ratio in laboratory conditions, where subjects are generally seated and asked not to move. The recommendation was to record EDA from the palmar surface of the non-dominant hand. This is not practical for 24/7 wear in today's studies. Moreover, there are new findings showing that the EDA can be meaningfully different on the left and right hand, and measuring on only the non-dominant side can miss some important information.

Empatica devices are optimized for ambulatory use in the real world where EDA from the wrist is much more practical. In practice, recording EDA from the wrist has a favorable signal-to-noise ratio in real-world conditions when people must use their hands.  During wake, EDA from the wrist will usually result in a lower signal than from the palm, although not always. Typically during sleep EDA will have a larger signal on the wrist than on the palm.

There are a number of articles describing how much the EDA on the wrist might differ from EDA captured from traditional recording sites on the palmer surface or fingers in the lab. Depending on the skin condition and the time the subject has worn the device, phasic signals from the wrist may be less pronounced (with a lower SCR magnitude) than those from the fingers or palmar surface.  The Empatica sensors are designed to have extra sensitivity to be better at picking up phasic responses from the wrist.

See this article published by SPR: Publication recommendations for electrodermal measurements

Dry electrodes depend on sweat as the electrolyte so it takes some time for a signal to appear under cold / very dry conditions (such as in an air conditioned laboratory, and with a participant who is sedentary and calm). Most of the time the sweat, that would evaporate under normal conditions, builds up when the electrodes are in continuous contact with the skin. Some protocols might require a warmup and baseline period to ensure that a conductive electrolyte is present between the skin surface and the electrode. The dry electrodes have the advantage of working comfortably for long durations where gels may degrade, get detached or cause irritation.  Gels worn for long times will change as they degrade over time, causing uncertainty in the data readings.

What is Baseline? How do I establish a baseline? For individuals? Multiple people?
Baseline is generally considered to be the average tonic level of an individual during rest conditions and in the absence of any discrete environmental event/external stimulus.  Baselines may be defined differently for different kinds of studies.
Standard practice – based on years of studies limited to labs, and usually for a lab session of less than an hour to gather data for scientific research papers – a typical practice is to have the person wear the EDA sensor for 15 minutes of rest prior to any tasks or measurements. During this period the subject should be asked to sit comfortably, hold still and relax. If they succeed in truly relaxing, then ideally the signal drops to be low and smooth, and once it does, then the lowest values are usually averaged to provide a baseline.
Best practice -  Empatica devices make long-term EDA measurement practical, so now you do not have to pretend that a 15 minute baseline in the lab is a true baseline.  Instead, you can get a 24/7 set of readings and look for naturally-occurring baseline data.  For example, measure what a person’s truly lowest tonic skin conductance levels are over daily life. This solves many of the problems that previously existed with short-term lab-only measurements of EDA baselines. The best way to get a baseline now is to measure long-term, and look for the lowest smooth period of skin conductance where there is no physical movement (where the accelerometer data is smooth), and where the temperature is clearly body temperature.

Non Responders - during your investigation you may encounter people who do not exhibit any phasic changes in EDA. A small portion of the general population do not exhibit significant changes in EDA in experimental settings. This can be due to a number of factors such as medications that suppress the sympathetic nervous system response.  Also, a person under anxiety may (surprisingly) have a very large signal on one wrist (usually the dominant wrist) and almost no signal on the other wrist.  If you aren't getting a large signal on one side, try the other side.


Know how to process the EDA signal

At this time Empatica does not provide algorithms to compute the tonic or the phasic features of EDA.  There are many algorithms and free software available online for computing tonic and phasic information from EDA, as well as other features of EDA.




Have more questions?