Z1.02.2Centre–periphery transitionsdesignresearch

Elevation to the centre when needed

Aliases: attentional pull · escalation · encalming

What it is

The full principle of calm technology is that information moves easily between periphery and centre. The periphery is the resting state, but both directions of travel must exist: user pull — when the user wants to know more, a near-zero-cost action reads the peripheral stream in detail; and system push — when the state changes enough to matter, a step in encoding intensity carries the information into the centre automatically.

Only with both pathways does "resting at the periphery" stop meaning "locked out at the periphery". A purely peripheral stream that cannot be read closely is decoration; users go open a screen to check, and the peripheral channel becomes redundant.

Why it happens

Each pathway exploits a different attentional mechanism:

  • Push rides on transient capture. Peripheral monitoring is insensitive to steady states and sensitive to abrupt ones: onset of motion, a luminance step, a new sound automatically pull attention toward the change. Design exploits this by encoding "a change worth noticing" as a step in encoding intensity — the lamp starts pulsing, a sound appears — so elevation happens by itself, without subscription or unlocking. Conversely, expressing importance as "permanently high intensity" fails: however strong, a steady state habituates; three days later nobody looks.
  • Pull rides on zero-learning-cost inspection. The entry point for reading closely must be near-instinctive: walk over and look, tap, glance at a companion surface. Every added layer (open the app, find the device, log in) makes pull less likely, and the credibility of the peripheral stream decays together with the ease of verification.

Neither track alone suffices. Pull without push leaves exceptional changes (a leak, an intrusion) waiting for the user to happen to ask; push without pull leaves the user's occasional "but how are things right now?" unanswered until the next push.

Studying it

Effectiveness of elevation is operationalised with paired indicators: hit rate (proportion of worthwhile changes noticed in time) and false-elevation rate (proportion of elevations irrelevant to the user's current activity), with main-task interference as the cost term.

  • McFarlane's comparison of interruption coordination (2002, TOCHI) systematically compared immediate, negotiated, scheduled and mediated strategies for bringing information to the centre. Key finding: negotiated interruption — where the user may defer — often outperformed immediate delivery on performance and stress, establishing that how elevation can be modulated is itself a design variable.
  • Change encoding studies: detection advantage of animation/transients over static presentation, and the speed at which the advantage habituates, are standard questions in ambient display evaluation.
  • Field log analysis: correlate elevation events with subsequent user actions (was there a response within X minutes?) to estimate real hit rates, as distinct from self-report.

One methodological caution: hit rate and false-elevation rate must be reported together. Reporting only hits rewards over-pushing and turns the system into an alarm; reporting only interference rewards silence and turns it into decoration. The trade-off between them is exactly where the product is tuned.

Where it stops holding

  • Periphery and centre are a budget. Frequent elevation collapses the whole calm structure into a notification stream; at that point the question is no longer "how to elevate" but "should there be this many streams at all".
  • Pull presupposes discoverability. A user who does not know a stream exists will never pull it. A peripheral display with no visible trace simply does not exist for a user who was never introduced to it.
  • Elevation is modality-bound. Visual transients fail when the user leaves the room; auditory elevation fails in noisy periods. Elevation of critical changes must take presence itself into account rather than assuming the user is always within sensing range.

Applying it

  • Encode importance as abruptness of change, not standing intensity: urgent = a still object starts moving or a new sound appears; ordinary = a brightness or colour-temperature shift within the periphery.
  • Give every peripheral stream a zero-cost inspection entry point: a detail display that lights up as you approach, a voice question answered by a knock — never "check by opening the phone".
  • Make elevation tiered: peripheral steady state → peripheral transient (captured if relevant) → centre-demanding (response required). Rules for crossing tiers must be explicit; no stream may jump from steady state straight to the centre.
  • How to check: instrument every elevation with the user's response or ignoring, hand-label relevance, and compute hit rate and false-elevation rate; demote the high-false-elevation streams one tier and observe again.

Related

  • Same group: Z1.02.1 Ambient information belongs at the periphery of attention · Z1.02.3 Forcing entry to the centre is an interruption
  • Nearby: Z1.01 The disappearance of computing · I1 Response time
  • Search terms: calm technology · attentional capture · notification · interruption

Cards in the same group

Quick Actions

Share

Share this page

ios_share

https://hci.top/en/handbook/Z1.02.2