Understanding the Body's Baseline Response
Physiological arousal is the measurable activation of the autonomic nervous system. It shows up as increased heart rate, skin conductance changes, respiration shifts, muscle tension, and pupil dilation. It is not an emotion itself, though people routinely confuse the two. When someone says they are "aroused" in a psychological context, they mean their body is in a heightened state of alertness or activation, regardless of whether that state feels pleasant or unpleasant. In practice, you measure it through a handful of standard psychophysiological channels. Skin conductance response, also called electrodermal activity, is probably the most sensitive indicator of sympathetic nervous system activation. Galvanic skin response sensors clip onto the fingertips or palm and track changes in sweat gland activity. Heart rate variability gives you another layer, though it is slower to change than GSR. Respiration belts around the chest pick up breathing pattern shifts. Pupilometry is more specialized but extremely informative for cognitive load and arousal mapping. I ran a study a few years back where we were trying to map emotional responses to stress-inducing video stimuli using consumer-grade biometric sensors. The problem was that one of our participants had a condition called hypohidrosis, meaning they produced almost no sweat. Their skin conductance readings were essentially flatline across every trial, making it impossible to use GSR as a feature. We had to fall back entirely on heart rate metrics and self-report, which gave us a much noisier dataset. It took three extra days of reprocessing to get anything usable out of that subject's data. The lesson was straightforward: always check your signal quality before you collect too much data, and have at least one backup channel ready.
Here is something most beginner guides skip. Physiological arousal does not tell you the valence of an experience. A spike in heart rate and sweat could mean fear, excitement, anger, or even intense concentration. The Schachter-Singer two-factor theory of emotion actually built on this ambiguity, arguing that arousal plus cognitive labeling produces the emotional experience you consciously feel. Without the cognitive component, you just have a body doing things. This is why self-report alongside biometric data is not optional, it is essential for interpreting what the arousal actually means. Another counter-intuitive point is that high arousal does not always mean high intensity of experience. Some people, particularly those with chronic anxiety or PTSD, run at a chronically elevated baseline. Their resting heart rate is higher, their skin conductance level is elevated even during supposedly calm periods. When you then measure a stimulus-induced change on top of that already-high baseline, the delta looks small, but the absolute state is already near-maximal. You can miss important individual differences if you only look at change scores rather than raw levels. Always report both. The tools themselves have gotten significantly cheaper over the last decade. Devices like the Empatica E4, Biopac systems, and even some newer consumer wearables can capture decent quality signals for research purposes. A basic setup with a GSR sensor and a chest-strap heart rate monitor will run you somewhere between three hundred and eight hundred dollars depending on whether you buy new or refurbished. Biopac goes well over two thousand but gives you laboratory-grade signal conditioning and software that saves you from spending hours cleaning raw data.
Common pitfalls to avoid:
Get the Full Details

- Motion artifact is the number one enemy. Any movement, even slight finger tapping, creates massive spikes in GSR that look nothing like true arousal responses. Use motion-tolerant sensors when possible and instruct participants to minimize movement during trials.
- Skin preparation matters more than people expect. Cleaning the sensor site with alcohol and lightly abrading the skin reduces impedance dramatically. Skipping this step can add noise that makes the data unusable without extensive filtering.
- Sampling rate needs to match your research question. If you are studying phasic responses to specific stimuli, you need at least 100 Hz sampling. Respiration and heart rate can be sampled lower, but mixing low-sampled channels with high-sampled ones complicates synchronization later.
The biggest limitation of relying solely on physiological arousal as a measure is that it is blunt. It tells you that something happened in the body but not precisely what, when, or why. For complex human experiences like grief, moral reasoning, or social dynamics, the mapping between body signals and subjective experience is loose at best. In those cases, combining arousal measures with qualitative interviews and behavioral coding produces far more interpretable results than any single channel ever will. If you are just starting out and want to experiment, I would suggest beginning with a simple Arduino-based GSR setup or a cheap Oura Ring or Apple Watch for heart rate data. Collect your own baseline readings over a week while doing normal activities, then introduce controlled stressors like timed arithmetic or public speaking tasks. You will quickly see individual variation that no textbook example prepared you for. Most people's arousal responses are highly context-dependent, and the first time you watch your own data, you tend to realize how much noise there is underneath the clean diagrams in textbooks.