Population Attributable Harm Estimation

Description

Estimates how much population-level UX harm is attributable to a specific pattern, combining exposure prevalence with relative risk to show what removing it would prevent.

The treatise asks 'how many users are exposed to this pattern across the ecosystem, and what is the attributable harm at population level?' Attributable fraction methods answer this: the population attributable fraction combines exposure prevalence with the relative risk of harm among exposed versus unexposed users, giving the share of harm that would disappear if exposure were removed. The treatise's arithmetic (1% unintended purchases across 50 million monthly users equals 500,000 harms a month) illustrates the scale logic. Rigour needs a credible relative risk estimate, ideally from experiments or well-controlled observational studies, and attention to confounding.

Attributes

Loss type
Unmigrated
Stage
Post-launch
Effort
Days

Steps

  1. Define harm outcome and exposure (pattern presence).
  2. Estimate exposure prevalence in the population.
  3. Estimate relative risk of harm given exposure from experiments or adjusted observational data.
  4. Compute attributable fraction and attributable cases with uncertainty intervals.
  5. Translate into preventable harm (money, hours, incidents) per period.
  6. Use estimates to prioritise interventions across patterns.

Sources

  • Lost Methods in UX — A Complete Treatise (Gopinath)
  • Levin ML (1953) The occurrence of lung cancer in man. Acta Unio Int Contra Cancrum 9
  • Rockhill B, Newman B & Weinberg C (1998) Use and misuse of population attributable fractions. AJPH 88(1)
  • Levin, The occurrence of lung cancer in man (1953), Acta Unio Internationalis Contra Cancrum
  • Rockhill, Newman & Weinberg, Use and misuse of population attributable fractions (1998), American Journal of Public Health

Also known as

  • Attributable fraction
  • Population attributable risk
  • Attributable harm
  • Population attributable fraction
  • PAF
  • Attributable risk
  • Attributable Fraction Analysis

Categories

  • Epidemiology & Population Health