Prediction of Self-Perceived Stress and Arousal Based on Electrodermal Activity – a paper by Tomppa Pakarinen, Julia Pietilä and Hannu Nieminen has been presented at the European Biomedical Conference EMBC in Berlin in July.
The researchers were keen to understand how individuals perceive stress and how devices can capture it. The need for this kind of study has been there for a long time. Today, prolonged stress is a common cause of work-related health problems and has major negative impact on employee wellbeing and productiveness. Being able to measure exposure to stress long term would provide a valuable tool for improving workplace and personal wellbeing and potentially reduce health-related problems.
In contrast to some earlier studies, the researchers in this study attempted to simulate actual work-related stress rather than induce extreme reactions.
Electrodermal activity (EDA) in evaluation of mental state
Questionnaires are commonly used to measure the individual’s subjective perception of stress. Physiological measurements are used for assessing the physiological responses related to stress and arousal. The most common measures are heart rate variability (HRV), electrodermal activity (EDA), heart rate, electroencephalography, respiration, and skin temperature. In previous research, HRV has been the most commonly used method.
Electrodermal activity (EDA) reflects the functions of the autonomic nervous system and was chosen to be applied in this study. It is often used for the evaluation of different mental states such as short and long-term stress. In the study, test subjects were exposed to a 3-phase test (relaxation, arousal, stress) during which EDA was recorded, and the self-perceived stress and arousal were assessed.
The results are promising for the use of EDA as a long-term measurement of work-related stress
In this study, the research team was able to reliably classify relaxation, arousal and stress-inducing phases of simulated work with high accuracy (94.1% with BIOPAC, 82.8% with Moodmetric smart ring), using a number of EDA features.
When comparing EDA to subjective questionnaires, the self-perceived stress and arousal were classified with much lower accuracy of 60.5–72.2%. Based on the results, it is possible that individuals are less able to recognize and interpret the level of stress they are experiencing in a particular situation than the measuring devices reading their EDA.
Overall the results are promising for the use of EDA as a long-term measurement of stress at work.
For Moodmetric the research is an important, continued validation of accuracy. The easy-to-use Moodmetric smart ring can provide information on the stressfulness of work-related situations almost as accurately as respective laboratory equipment designed to measure EDA.
The Moodmetric smart ring measures stress in real time. It is a user-friendly way to capture electrodermal activity (EDA), which tells about the activation of the sympathetic nervous system.
This series of articles is about stress and the ways to measure it. Parts 1 and 2 describe the fight or flight reaction and how the body reacts to chronic stress. Methods for long-term monitoring are presented in part 3. In this article we explain the Moodmetric measurement and how to interpret the data.
How is the Moodmetric ring measurement done?
The Moodmetric smart ring measures stress as a physiological phenomenon. This is done by interpreting electrodermal activity that tells about the sympathetic nervous system activation.
Increased sympathetic nervous system activation makes sweat glands work harder. This can be seen in greater conductance of the skin.
Conductance measurement requires two electrodes of conductive material. The band of the Moodmetric ring works as a set of electrodes. It consists of two silver coated steel rings and an insulator band in between them.
To achieve a good and steady reading, the electrodes need be in contact with an area of the skin where the eccrine sweat gland density is high. This density can vary from 400/cm2 on the palm of the hand to about 80/cm2 on the upper arm.
Accurate skin conductance measurement can be done from the palm side of hand. A finger ring has an optimal position from the accuracy point of view and the Moodmetric ring is a ring for this reason. The actual point of measurement is on the inside of the finger and at best the signal is completely unbroken.
The ring measures continuously and stores the data inside the ‘stone’. The mobile app does not need to be turned on or open, nor the phone near the ring. When the app is activated and the calendar icon chosen, the ring sends the data to the app via a bluetooth connection. The measuring and data storing to the ring continue immediately. The real-time signal – in other words, the ongoing measuring in process – can be observed at any time on the app.
The Moodmetric signal is the violet curve on the mobile app that can be observed in real time. It is the raw measurement signal, autoscaled in order to have the whole amplitude visible even during strong reactions.
The curve enables the analysis of single reactions. Even a thought – excitement, idea, awe – can cause a peak and is registered with only a 1-2 second physiological delay.
Interpreting the raw signal requires expertise in the measurement method and understanding of the possible sources of error.
The Moodmetric stress measurement data
The raw signal for electrodermal activity (EDA) is difficult to interpret. Strong reactions can easily be spotted on the curve, but mathematical methods are needed to gain further insight.
The Moodmetric level has been developed to provide accurate EDA measurement data that is easy to interpret. The algorithms count an index from 1 to 100 so that 1 is the lowest the person can reach. This is possible, for example, in deep sleep. 100 is the maximum level of arousal, strongest possible reaction. Since we are all individuals, the Moodmetric measurement method is designed to find the minimum and maximum levels of each person within the first 12 hours of taking the ring into use. This is called the calibration period. The ring should be recalibrated when handing it over to another user.
The MM level on the app, both the numeric value and the curve being drawn, show the person’s ongoing activity and level of excitement for the past few minutes. The MM level does not indicate single reactions, but changes fast if the arousal level of the person increases or decreases rapidly. Increasing is first visible in the growing amplitude and the raising trend of the Moodmetric curve, then in the higher numeric value for the MM level.
It is easy to get to 100. The challenge lies in working out the ways in which to get the MM level as low as possible.
When calculating the MM level, the algorithms minimize the effect of finger movement and skin moisture. The MM level is comparable between users. If two persons are placed in the same environment and situation, it is possible to observe which one is calmer.
There is no momentary optimal value
It is normal for the Moodmetric level to fluctuate between 1 and 100 during the day. No momentary value is good or bad.
It is equally important to acknowledge that the Moodmetric measurement does not tell whether a reaction is positive or negative. The Moodmetric ring is not a detector of emotions as such.
The fluctuation of stress levels is different for everyone
There are people who react fast and strong, while others respond more calmly. For instance, creative people are often very susceptible to stimuli, which can translate into high and fluctuating MM levels. In contrast, a person doing work that requires much concentration over long periods of time might have low MM levels throughout the working day.
The levels can be high due to excitement and energy, or low due to intense concentration. All this is positive. The levels might also be high due to pressure and lack of control at work. Low figures in turn can tell of boredom or even depression. For an accurate assessment, the measurement data always needs to be complemented with the ring wearer´s own perception of the situation. Only the person in question knows whether a set 100 tells about excitement or irritation.
A successful team is heterogeneous also by the way they react
On average, it is productive at work to have people of different temperaments. Depending on the work, of course, a team consisting only of individuals with either a very mild or a very strong natural way of reacting is probably not as productive as a more heterogeneous team.
The Moodmetric stress measurement data increases our understanding of the different ways in which people react to various situations. This is important from the point of view of health and well-being; to challenge and grow, we should look within and learn from our experiences rather than compare ourselves to others.
The 24-hour MM average level is what it’s all about
The Moodmetric smart ring measures the reactions of the sympathetic nervous system on a scale of 1 to 100. When the 24-hour average value is around 50, there is enough recovery in correlation to the load.
Fluctuation of the MM levels during the day can vary much between different individuals. High daily figures are no cause for worry if the person feels energetic enough and has a restful night: It is the 24-hour MM average level that counts. The same goes for someone whose MM levels don’t peak during the day.
The point is demonstrated in the graphs below: Two persons have exactly the same 24-hour MM average level of 46, which signifies a good balance between rest and activity, but their autonomous nervous systems are activated in different ways and by different stimuli during night and day.
The more you use the ring, the more information you gather to help you adjust your behavior. It is therefore recommended that the ring is worn also during nighttime.
Feedback from users of the Moodmetric ring indicate that the MM level correlates with a user’s own perception of their situation. If the 24-hour average is around 50, the energy levels are normal. Values over 50 for long periods of time go hand in hand with the feeling of being overloaded; cracks in one’s well-being begin to show. Very low daily figures, on the other hand, might be a sign of depression.
The Moodmetric measurement helps the user to gain balance between load and recovery. The sources of stress and ways to recover differ from person to person. For this reason, the measurement data should never be examined in isolation, but further self-assessment by the individual is required. The Moodmetric ring and app are tools for a person seeking more balance in their life.
The complete set of 5 articles explains the Moodmetric measurement, science behind and the applications:
The Moodmetric data analytics tool is created to support research and development projects related to EDA (electrodermal activity) measurements.
The tool enables researchers and developers to quickly process and visualize large Moodmetric data sets in uniform manner. It generates both group and individual level reports based on input data from wearable devices. The tool is released as open source for anyone to benefit of the Moodmetric measurement data in various use cases.
Download the Moodmetric data visualization tool instruction here.
I have used Moodmetric ring on a regular basis for almost two years now.
When I first got to know Moodmetric, my life was not terribly hectic, I did not feel to be particularly stressed. I did not check my stress levels from the Moodmetric app continuously, as there was not much change – figures were quite low, nothing to be alarmed of.
Until last January.
It was already during Christmas break, when I started to notice the first signs. The feeling of hopelessness and a hint of bitterness had started slowly fester inside me. Two and a half years of burdening family life with extreme efficiency had resulted in complete loss of energy. Our family size had undergone significant change few years back, as the number of our kids went from one to three at once. Without a proper safety net, my husband and I had worked like machines to take care of our family. After 2,5 yrs. we were both exhausted and the only thing keeping us sane was being able to go to work to our paid jobs. Yes, it is a bit twisted, that you go to work to recover from family life.
At work my colleagues would have their daily laughs, because I started forgetting things. I would even go to a meeting and be very impressed by a work very nicely done just to hear that I had been part of it. Memory problems showed up.
By the end of January I noticed that I couldn’t pull through my regular kettlebell exercises anymore let alone improving the performance. I was just tired. I wanted to exercise, but I was too tired to get to it.
Then I got my first flu. And a second flu. And a third flu. Soon I noticed that I was the only one in our family getting sick all the time. (By the end of July I had been ill almost ten times.) My immune system had failed me.
In March I decided to take a personal risk as I threw in my lot with the Moodmetric team. I became a co-owner in the company. Becoming an entrepreneur was something I had always dreamt of, but I had little doubts about the timing. It was a great move for me, but I knew that the positive stress – excitement – could be a challenging combination with the cumulated stress from the family circus. Would I be able to unwind from all the positive stress ahead me?
In April it became obvious to me that my overall stress levels were higher than usual. Wearing the Moodmetric ring, I had data to back this up – I could witness the change in figures from the app. I was alerted, but not afraid, because I am a very good sleeper and felt that I got the recovery I need. In overall I was optimistic about the future.
The spring 2016 and early summer were hectic and exciting. We were rewriting the company vision and strategy, and talking to loads of people with to get feedback. What we heard was so supporting that at times I couldn’t restrain myself from working unreasonable hours. On the other hand, I had started to question my work performance and felt like I was not working enough. I knew I had gone into overdrive a big time, but was hoping that I could make it to my summer vacation.
When July and summer vacation started,I breathed freely again – I had made it to the safety zone and now I could unwind and recover!
Only, that feeling lasted for a short moment, because sleep disorders kicked in. I started having problems with falling asleep and my heart would race for anxiety. Bedtime became one of my least favorite time of the day, because I was afraid I couldn’t fall asleep.
Needless to say, summer vacation came in too late. I had neglected my recovery and crossed the line that I didn’t wish to. I am well aware that one shouldn’t play with sleep disorders. My Moodmetric daily diagrams were screaming red. That is when I decided to start adding more unwinding moments to my everyday life and not just wait for the next vacation. I did not want to welcome chronic stress into my life.
In short, the lifestyle and behavior changes I adopted are mindfulness exercises, less heavy exercise and replacing with long walks, adding micronutrients, prioritizing sleep, etc. I will write down how I tackled chronic stress in very much detail in my next blog, due out soon!
With any sympathetic nervous system activation, skin reacts and becomes a better conductor of electricity. This can result from emotional, cognitive or other psychological origin. The phenomenon is known as electrodermal activity (EDA) and it is widely adopted in psychological research (1). Other commonly used terms for this phenomenon are skin conductance response and galvanic skin response.
EDA is generated by the activity of the sweat glands. Moodmetric measures the palmar skin on your finger. The palmar skin is the recommended EDA measurement location, since it has the highest eccrine sweat gland density (2). You can measure EDA elsewhere as well but the reliability is not as high or as easily achieved.
The unconscious actions of the human body are regulated by the autonomic nervous system. It consists of the sympathetic and parasympathetic part. The parasympathetic part controls the body’s rest-and-digest functions and the sympathetic part controls the fight-or-flight reactions. By examining electrodermal activity we can understand the sympathetic nervous system reactions.
The sweat glands are exclusively innervated by the sympathetic nervous system. This makes EDA an ideal measure for sympathetic activation (2). Electrodermal activity correlates to cognitive and emotional arousal, and high responses are caused by e.g. stress, enthusiasm, anxiety, joy, anger (1, 3).
Mobile EDA devices have been used by scientists for some time (2, 4). The Moodmetric ring is an unobtrusive option to follow EDA responses real-time and in long term.
The signal accuracy has been proven in a study of 24 people by the Finnish Institute of Occupational Health by J. Torniainen et al.. The accuracy against a laboratory grade reference was found to be 83 %. They conclude:
“Clearly the ring sensor can be used to measure a valid EDA signal as indicated by the similarity of both event-related responses and the calculated features. The accuracy of the Moodmetric EDA Ring is adequate for psychological and physiological research when weighted against the advantage of conducting ecologically valid experiments outside laboratory conditions.”
The results were accepted for publication in the 2015 conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2015, August 25-29).
The Moodmetric measurement accuracy has also been studied at the University of Tampere, Finland in 2014. The correlation with the reference equipment Nexus-10-MK II was good. Comparison details in charts below.
With skin conductance level (SCL) we refer here to raw measured skin conductance without any filtering. The figures below show the comparison of the two devices.
Further reading in an article by Jari Torniainen and Benjamin Cowley, published in August 2016:
The Moodmetric ring and app are the simplest solution to measure stress and excitement. Take the benefit of the Moodmetric continuous measurement and real-time feedback to support your research.
The Moodmetric ring is developed to analyze the activation of the sympathetic nervous system, ie. the fight-or-flight reaction. This can be detected through the skin conductance response measurement. It is best done from the palmar skin or fingers, which makes the ring form perfect to obtain accurate data.
Where can the electrodermal activity measurement be applied
This phenomen of electrodermal activity / galvanic skin response has been an input for research for over hundred years. It is used in large extent for instance in
– Psychological research
– Clinical research & psychotherapy
– Media & advertising testing
– Consumer neuroscience & marketing
– Usability testing & UX design
How the Moodmetric ring is used
The Moodmetric ring is a non-intrusive way to measure the electrodermal activity, which enables long term measurement outside a laboratory setting. The ring is worn as a normal ring. It connects to a smartphone app via bluetooth smart and real-time measurement works within 5m radius.
The app shows the electrodermal responses measured by the ring and the output of the signal analysis which is the Moodmetric index (0…100). In addition, the Moodmetric number among other data, so the ring can be worn without phone connection. The data can be synced any time.
Direct data streaming from the ring to Windows PC is possible with a BLE dongle. This enables also access to EDA raw data.
In addition to EDA sensors, there is a three axis accelerometer inside the Moodmetric ring. This enables for example measuring the step count, that is included in the app.
The measuring and output
Due to the advanced signal analysis methods to derive the Moodmetric index, the measurement works for everyone. It adapts to the typical level of the person using the ring. The ring learns the levels of the user, becomes personal, and always gives a number between 0 and 100, with a typical activation level of 50. If the user changes you can reset and the learning starts from the beginning.
The algorithm also includes an advanced artefact rejection. The electrodes touch the skin and even if your hand moves the algorithm rejects the disturbances in the signal caused by the electrode movement.
The Moodmetric measurement accuracy and current use in research
The Moodmetric ring feasibility for research has been verified by the Finnish Institute of Occupational health. Link to the EMBC ´15 conference paper.
It has been used in following research programs:
Pop up, 2015-2016
‘Pop up – knowledge work productivity’ research project provides new research results, practical methods and measurement tools for developing knowledge work productivity and well-being at work. Knowledge work is analyzed through a work system including physical, virtual, social and emotional environments. This project develops and utilizes participatory Pop up –method for designing and testing new work environments and practices that provides more productive ways of working. In addition, the project develops metrics and measurement tools for analyzing the impacts of new work designs. Mobile devices, sensors and applications are utilized to study fluency and experienced well-being and productivity of knowledge work.
Project’s multidisciplinary team comprises researchers from Tampere University of Technology (TUT) and Aalto University. The project is carried out in close collaboration with industry partners: Tampere Region Economic Development Agency Tredea, Arkkitehtitoimisto Helamaa&Heiskanen, University Properties of Finland Ltd, Martela and Moodmetric.
TUT Novi research center: Maiju Vuolle, maiju.vuolle(at)tut.fi
TUT Human-centered technology: Kaisa Väänänen, kaisa.vaananen(at)tut.fi
TUT School of Architecture: Jenni Poutanen, jenni.poutanen(at)tut.fi
Aalto Virtual and mobile work research unit: Matti Vartiainen, matti.vartiainen(at)aalto.fi
DEEVA, 2016 -2019
DEEVA project utilizes the opportunities of digitalization to create value from data and to develop new, customer driven service products and methods which support value co-creation and that are based on deep understanding of customer experience.
The research question is as follows: what kinds of means, modes and contexts combining data, emotions and experiences digitalization enables. The project is carried out with a large and versatile network of enterprises. The participating 20 companies vary in size and industry, e.g. energy, media, bank, ICT, real estate, commercial and service sectors are represented in the project. Multisectoral group of enterprises enhance both co-learning and gaining new insights into the research topic.
The project is executed by a multidisciplinary research consortium of three universities: Tampere University of Technology, Turku University of Applied Sciences and Tampere University of Applied Sciences in co-operation with six international universities. In addition to knowledge and publications for different target groups the project will create tools and applications for measuring customer experience and analyzing emotion data in real-time. The information provided by the tools and applications can be used in everyday activities of companies to support e.g. management of multi-channeled service environment and development of new service products and co-creative ecosystems.
Is measuring workplace well-being a plot by the employer? Can the employee benefit of it? What is worth tracking and what do the results tell? If I take 10000 steps per day, sleep 8 hours per night and eat healthy meals at regular intervals, will I sell more cars than my colleague car salesman?
At the end of the day the person with better sales skills might still get bigger bonuses. On the other hand, I might feel better and spend less days off sick. I might even feel energetic despite hectic workdays and take setbacks like loosing a deal lighter.
Should workplace measuring only take place in the workplace?
Activity trackers and other wearables are already part of our lives. They were initially intended for personal use with goals set by the wearer. A sleep monitor can hardly be a tracker used in workplace and collecting thousands of steps in the office is only possible with the running mat under the desk.
These devices can be used to monitor general well-being and they can potentially give valuable information. They can also contribute to better work performance if the person is a motivated wearer. Someone else (the boss) reading the results might still not have the full picture. Only the measured person can put the results into context and know if today is a good day to make great results.
Pulse, heart rate variability, electrodermal activity (EDA) /skin conductance and EEG measurements can be done at the office, while sitting down. Heart rate signal however needs processing before it tells about stress levels. EDA and EEG give instant emotional load results which are not tied to physical reactions. EDA is best measured from palmar skin, EEG requires a headband with at minimum three contact points to forehead.
The workplace well-being measurement might be best done in the workplace and only if the employee thinks it is a good idea. Tracking outside the office could be part of a wellness program with limited length, on a voluntary basis.
Having the employer know when I go to sleep or how many times I was waked up by my kids – maybe not. But being able to show which meetings are the most stressful or how calming it is to take a 5 min walking break outside, why not!
This article explains the difference of using a wristband or a ring to measure your emotional intensity levels.
Many things can get you emotionally activated. A close by situation with a colliding car can certainly light all the inner alarm systems, but milder things can also cause strong emotional reactions: seeing your children after a school day, a victory of you favourite basketball team or forgetting to buy paper for your printer.
With any emotional activation, your skin reacts and becomes a better conductor of electricity. This can be resulting from emotional or other psychological causes. The phenomenon is known as the skin conductance response or electrodermal activity (EDA). You may also run into an older term called galvanic skin response.
Based on the scientific research there are only a few places in your body where the EDA can be measured accurately and easily: the palms and the soles of your feet. These are the places where human body has the highest density of eccrine sweat glands that response to the emotional stimuli. You can measure the EDA elsewhere as well but the reliability is not as high or as easily achieved.
We want to bring to the market a very reliable and accurate measurement of emotional activation and intensity that is available for everybody. The choice for having a beautiful jewelry ring as the measurement form was an easy one.
It provides the best optimal measurement data that can be detected from the palm side of wearer’s hand. A ring is also a natural thing to wear and can be used daily as any other jewelry. It is easy to forget that this small, beautiful and non-intrusive jewelry detects your emotional levels and helps you in your life with its data.