A signal is not an explanation
Biofeedback begins with a measurable physiological signal such as heart rate, breathing, muscle activity or skin conductance. The device measures a signal; the software then displays it in a form the user can observe. Interpretation comes later.
This distinction matters because many wellness applications use the word “biofeedback” for any before/after score. A self-rated calmness value is useful data, but it is not the same thing as a physiological sensor measurement.
Three layers of feedback
Measured signal: data from a sensor. Subjective report: what the user says they feel. Interpretation: a conclusion drawn from one or both. Mixing the three creates false precision.
Subjective ratings still have value
Subjective data is not worthless. Pain, calmness, clarity, tension and perceived energy are experiences that only the person can directly report. The key is to use the same question and scale repeatedly, timestamp it and avoid presenting the score as a medical measurement.
What makes a feedback loop useful?
A closed feedback loop lets the user observe a signal or state and then deliberately change something—breathing, attention, posture or session protocol—before observing again. The process becomes more informative when conditions are reasonably consistent and the same variables are tracked over time.
Before/after is not enough for causality
If calmness is higher after a session, the change is real as a self-report. The cause is not automatically known. Time, expectation, breathing, rest, environment or the symbolic ritual may all contribute. A responsible application reports the change without overclaiming the mechanism.
First define what we are actually trying to observe
Feedback is useful only when the target is defined clearly enough. If we want to track relaxation, we can specify a physiological signal, a subjective rating and a time window in advance. If we want to track attention, we need a different outcome. Combining many goals into one generic “wellness” number can hide what actually changed.
A simple N-of-1 approach
For personal exploration, a repeated single-person design can be useful. Record a baseline over several days or sessions, introduce one change, and continue measuring under similar conditions. If the signal changes after the intervention and the pattern repeats, that is more informative than a single before/after capture.
This approach still does not guarantee causality, but it reduces the chance of treating one unusual day as proof of an effect.
Artefact, expectation and genuine change
Any feedback system can be influenced by sensor quality, capture protocol and the user's expectations. Keeping raw data, or at least enough detail to review recording quality, helps distinguish a changed signal from a changed measurement process.
A subjective improvement can be meaningful in its own right, but it is better labelled as a subjective report rather than presented as a physiological measurement. The two can coexist without being confused.
A well-designed system also shows uncertainty
Professional feedback should not display only a result. It should also expose capture quality, stability and relevant limitations. Messages such as “poor signal”, “unstable recording” or “insufficient repetitions” are not weaknesses. They show that the system distinguishes data from interpretation.
Three questions before interpreting a graph
Before assigning meaning to a graph, ask three simple questions. First: is the signal technically good enough? Second: is the change larger than the person's usual day-to-day variability? Third: do we have enough repetitions to avoid explaining one unusual event?
These questions often improve interpretation more than another colour scale or another composite score.
Why display design matters
An interface can push the user toward a conclusion before the data justify it. A large red number feels alarming even when it represents a small deviation with low confidence. A better display includes the value, reference context, personal trend, signal quality and, where possible, an indication of uncertainty.
Research-oriented tools also benefit from access to raw or intermediate data. The result then becomes a traceable process from capture to interpretation rather than just a score.
How PICALLW separates the layers
In Digital Radionics, subjective ratings and notes are treated as reflective data. Where a real sensor is used elsewhere in the PICALLW ecosystem—such as HRV—its physiological status should be labelled separately. An AI summary may combine information, but it should never turn an interpretation into a measurement.