Q3.10.1Arousal–valence dissociationdesignresearch

Physiological signals index arousal, not valence

Aliases: GSR not emotion · skin conductance · affect circumplex

What it is

Skin conductance, heart rate, and pupil size track autonomic arousal: how strongly the body is mobilized. Valence is whether the experience leans pleasant or unpleasant. The same rising electrodermal wave can be the thrill of clearing a level or the jolt of an error. Writing a physiological rise as “negative emotion” or “liking” treats a one-dimensional intensity as a signed feeling.

Why it happens

Sympathetic activation increases sweat-gland activity, heart rate, and pupil size for threat and for reward alike. Peripheral signals sit far from the appraisal that produces valence, and they drop the sign “is this good or bad for me.” International affective pictures can manipulate arousal and valence separately; interface tasks often bind them: failure is both more awake and more negative, success both more awake and more positive, so an arousal gap is narrated as a liking gap. Without independent valence evidence—facial EMG, an explicit pleasant–unpleasant report, or a task outcome that carries a sign—a physiological curve cannot testify to affective direction.

Studying it

Collect arousal and valence on separate tracks. Valence can be corrugator/zygomatic EMG, a pleasant–unpleasant scale, or a task whose outcome has a sign (win/lose, accept/reject). Report whether arousal changed, then ask direction separately; do not treat an electrodermal peak as “stress” or “delight.” Induction materials should include high-arousal positive, high-arousal negative, and low-arousal controls, to show that the index is insensitive to valence. Claims that an interface “made people more anxious” need valence-side evidence; otherwise the sentence is only “autonomic activation was higher.”

Where it stops holding

When the task already defines direction—explicit punishment or reward—arousal can measure intensity, with direction read from the task rather than from the body. Some central measures and facial EMG are more valence-sensitive, but their conclusions cannot be gifted to skin conductance. Drugs, caffeine, and clinical states shift arousal baselines. Marketing copy that treats “a response” as “an emotional connection” is rhetoric, not measurement.

Applying it

  • Title physiological plots as “arousal change,” never as “good or bad emotion” or “users were happy/annoyed.”
  • If the decision needs a sign, add a valence measure or an outcome with a clear win/lose, and read the two columns together.
  • For features that can thrill or frustrate (gamification, timed offers), do not let skin conductance decide whether to ship.
  • Default external wording to “autonomic activation rose; valence was not measured” until valence data exist.

Related

  • Same group: Q3.10.2 Motion and environment contaminate the signal · Q3.10.3 Baselines and within-person comparison are required
  • Adjacent: Q3.09 Eye-tracking experiments · Q3.06 Task success rate
  • Search terms: arousal–valence dissociation · electrodermal activity · affect circumplex

Cards in the same group

Quick Actions

Share

Share this page

ios_share

https://hci.top/en/handbook/Q3.10.1