Abstract

People track data to know themselves — not only to hit a number. Li, Dey, and Forlizzi map five stages from choosing what to measure to acting on what you learned, and show why a crack in an early stage stops the rest.

Li, I., Dey, A. K., & Forlizzi, J. (2010). A stage-based model of personal informatics systems. In Proceedings of CHI 2010 (pp. 557–566). ACM. DOI

Key findings

  • Five stagesPrepare (what and why) → collectintegrate (one view) → reflect (what it means) → act (one change).
  • Early breaks cascade — Bad aim, spotty logging, or siloed apps mean reflection and action never get fuel — a prettier dashboard does not fix empty capture.
  • Loop, not ladder — You revisit stages; design should allow going back without shame.
  • Where apps fail — Most tools love collection and under-serve reflection and action — where your life actually moves.

Limits

Built from interviews and surveys (2010) — a map of common patterns, not proof that one app works. Tools changed; the stage names still help you diagnose your own stack.

Application

When you log voice notes or metrics but nothing shifts, walk the five stages once: unclear aim · missed capture · data stuck in silos · no review moment · no small committed move. Data selfie · Weekly Review.

Adjacent