Against matched controls, year-2 English Wikipedia views rose about 60% after deaths before age 60 and fell about 11% after deaths at 85+ (1,276 deaths, 2016–2024)
Question: does death increase long-term attention to a person's English Wikipedia article? Prior studies compare views after death with the same person's views before death. We compared each person with matched controls in the same calendar windows. Data: a 6% hash sample (7,488) of all 123,455 Wikidata humans with an English article who died from August 2016 to September 2026, 23,925 living controls stratified by birth decade, and daily user pageviews from July 2015 to September 2026. Treated: 1,276 people who died 2016–2024 with a day-precision date and at least 5 views per day in days −395 to −31. Each had 5 controls (living or not yet dead), matched exactly on gender and on birth year within 2 years, and by nearest neighbour on the level and trend of pre-death views. Results: relative to controls, the excess views decayed and reached about zero 16 months after death. The mean year-2 (days 365–729) effect on log(mean views+1) was +0.018 (95% CI −0.011 to 0.047). Years 3–8 were also near zero. Three placebo runs with pseudo-dates for living people gave −0.016 to −0.032, so effects below about ±0.05 are not interpretable. Age at death set the sign: +0.467 (0.318 to 0.623) before age 60, +0.100 at 60–74, +0.010 at 75–84, and −0.115 (−0.150 to −0.080) at 85 or older. Placebo runs showed no age gradient. The gradient held with living-only or later-death-only controls, 10 controls, other calipers and thresholds, and step filters. It also held in an adjusted regression (−0.011 per year of age), which also gave a large effect of unnatural death (+0.70). People with 15 or more language editions had a +0.133 year-2 effect. Limits: we counted views only for the current title, the bootstrap ignores control reuse, living status comes from missing Wikidata death dates, and only articles with pre-death traffic were analysed.
Wikipedia pageviews · collective attention · collective memory · death · matched controls · event study · Wikidata · placebo test · age at death · attention decay
References
2- doi:10.1073/pnas.2106152118
Postmortem attention in news and Twitter with a self-baseline design and power-law decay; we test the long-term part with matched controls on Wikipedia.
- doi:10.1038/s41562-018-0474-5
Model of the decay of collective attention and memory that uses Wikipedia views.
Cited by
3 publicationsPublic catalog entries on Sciscape.
- Against matched living controls, English Wikipedia views after 17,866 deaths in 2019–2023 were 8% higher in year 2 and not higher in year 4; year-4 change was +40% below age 60 and −11% at 85+
[Legacy relation: reproduces] Same question and a similar matched design on a 6% sample; we reproduce its age-at-death gradient with all 2019–2023 deaths, page-id counts from bulk dumps and placebo calibration, and we extend the follow-up to 7 years.
- supportsMatched controls show no median year-2 rise in English Wikipedia views after 1,026 sampled deaths (2017–2024), a fall of about 10% per decade of age at death, and inflated one-year plateau fits
Independent replication with a different random sample, sampling frame, traffic threshold and code; we find the same near-zero year-2 effect and almost the same age-at-death gradient (-0.047 log10 units per decade).
- Death gave a 2.4-fold year-1 rise in English Wikipedia views but no clear median change in years 2–4 against matched living people (1,517 deaths, 2016–2021); effects fell with age
[Legacy relation: reproduces] Independent matched-control study of the same question on a different sample; we replicate its null average year-2 effect and its age-at-death gradient, and we report a difference in years 5-9.