Trump Posts & Context — an observational co-occurrence timeline
A neutral timeline placing Donald Trump's social-media posts alongside
contextual inputs (Fox News web-coverage keyword volume, S&P 500 shocks,
court events in his own litigation) and policy outputs (presidential
documents). It shows what occurred near in time to what — nothing more.
What this is, and what it is not
Observational co-occurrence, not causation. Things
appearing close together in time are not evidence that one
caused the other. This tool measures no cause, intent, or psychology,
and is not a measure of anyone's "state of mind".
The Fox ↔ Trump relationship is a feedback loop.
Web coverage may follow his posts as easily as precede them, or both may
follow a third event. Direction of influence cannot be inferred
from timing alone.
“Fox News coverage (web)” counts WEB coverage, not TV airtime.
This track measures the daily volume of foxnews.com web coverage
mentioning each keyword (GDELT DOC 2.0 “Volume Intensity”) — not
what was spoken on air. The earlier TV-caption source remains
gated at 2024-10-10 and is not used here; this is a separate web
proxy.
The court track is context, not a claim. It marks
RECAP dockets in which Donald Trump is a named party (CourtListener,
anonymous), placed at each docket's filing date. Many are official-capacity
suits naming the President. Court dates are date-only, so this lane is
excluded from the epoch analysis below; it is display/context only.
Scheduled economic releases are a curated calendar. The
econ lane marks when CPI, the Employment Situation, PPI, GDP and the FOMC
statement were released, at their fixed Eastern times (DST-aware). They are
shown as context; their placement asserts nothing causal.
The epoch analysis tests temporal association only. Its
window and metric are pre-registered and fixed; it reports a
permutation p-value per event type. Several types are tested
(multiple comparisons), and small samples mean low power — so a null or
inconclusive result is the expected, honest outcome, never proof of a
cause or reaction. See the section below.
Settings in use:loading…
All timestamps are UTC (ISO-8601). The axis and every
tooltip below are shown in UTC. Eastern-time sources (markets, Fox) were
converted with daylight-saving handled correctly.
One deliberate exception: the posting-rhythm punchcard
alone buckets hour-of-day in US Eastern (America/New_York,
DST-aware), because time-of-day is only interpretable in local time. Every
other view and all tooltips remain UTC.
Posting tempo — Truth Social posts per day
One bar per day = number of posts that day (bucketed in UTC, to
match the timeline axis below). Markers overlay market shocks (inputs),
scheduled economic releases (inputs, top edge) and presidential documents
(outputs) on the same time axis. Descriptive volume only.
Truth Social post Fox News coverage (web) — keyword volume (daily) Market shock ▲ Market shock ▼ Court (Trump litigation) — docket filed Presidential document Economic release (scheduled) Reaction window (visual only)
Posting rhythm — posts by weekday and hour of day
Each cell = total posts in that weekday/hour slot across the whole
term, shaded by count. Hour-of-day is bucketed in US Eastern
(America/New_York, DST-aware) because time-of-day is only interpretable locally.
Descriptive rhythm only — not a measure of state of mind.
Epoch analysis — posting in the hours after an event vs. a random baseline
For each event type, this compares the number of Trump posts in a
pre-registered, fixed window of 0 to +6 hours after each event against a
baseline in which the same number of events are scattered at random across the term
(5,000 permutations). Computed in Python in the pipeline; this page only draws the result.
loading…
Prospective test: jobs-report posting
A single hypothesis, pre-registered on 2026-06-09
(see PREREGISTRATION.md) and computed in Python. Unlike the
exploratory analysis above, this one test was fixed in advance, so it is not
subject to the multiple-comparisons problem — but it only becomes conclusive
once enough future releases have accumulated.