Every Monday we pull the week's most-covered stories out of Signal/noise's live source feed and look at the same event through the outlets that reported it. Not to score who was right, but to show how the choice of verb, detail, and emphasis quietly steers what a reader takes away. This week's set — 15 stories drawn from across the political spectrum on October 26 — gave us four worth walking through side by side.
Brazil's runoff: a result, or a referendum on Trump?
The most-covered story in this week's set was Brazil, where Luiz Inácio Lula da Silva won the presidential runoff and the losing camp refused to concede. Signal/noise logged 23 sources on it, spread fairly wide: five left, eight center-left, three center, six center-right, and one right-leaning outlet.
The framing split showed up less in the facts than in the frame around them. On the left, Mother Jones — in partnership with Agência Pública — ran the result under the headline "Lula's Surprise Win Blocks Trump's Push for a Far-Right Brazil," reporting that Washington had spent the two weeks before the vote trying to pull Latin America rightward. The New Republic called the outcome "a win for the rule of law and a blow to Trump." In both, a Brazilian election is primarily a story about the United States.
Center-right and right outlets covered the result too, but Signal/noise did not capture an equivalent frame from them tying the win to Trump — which is itself telling. When one side reads a foreign election as a domestic proxy fight and the other largely reports it as a foreign election, readers of each come away with different stories bearing the same dateline. You can see the full source spread on the Brazil election story page.
Serbia: "a seismic political shift" or "a brief concession notice"?
Serbia offered the week's cleanest framing contrast, and it lived almost entirely in the adjectives. The same event — former president Aleksandar Vučić conceding defeat to a student-led movement — ran at CNN as "Serbia's Aleksandar Vucic concedes defeat to the student-led movement in a seismic political shift" and at the Financial Times as "Serbia's ex-president Vučić concedes defeat to Student List."
One headline frames the result as a political earthquake; the other files it as procedural news. Neither is wrong. But "seismic political shift" and a plain concession notice are two different invitations to the reader — one to feel the stakes, one to note the fact and move on. This is episodic versus thematic framing playing out in a single word.
There was also a quieter pattern in the 10 sources covering it: three left, five center-left, one center, one center-right, and zero right-leaning outlets. The Financial Times sat at the right edge of a set whose right flank was otherwise empty. When an entire side of the spectrum skips a story, that absence shapes perception as much as any headline does — the mechanism we cover in bias by omission.
Venezuela: a homecoming, told as a reception
Venezuela's exiled opposition leader María Corina Machado returned to Caracas, and the outlets covering it — 10 in the set, again weighted left and center-left with no right-leaning sources — largely told it as a reception story. Al Jazeera, the BBC, and DW framed the return around the crowd: thousands in the streets, a hero's welcome, a dramatic homecoming after what the BBC specified was "more than 10 months" in exile and a speedboat escape.
That is a legitimate frame, and a vivid one. It is also a choice. A homecoming framed around the size of the welcome emphasizes popular legitimacy; the same facts framed around the government she returned to confront, or the risk she took in returning, would read as a different story. The Venezuela story page lists the outlets so you can compare openings yourself.
The 2026 midterms: the same race, three different entry points
On the US midterms, the clearest captured frame came from the right. The Washington Examiner ran three pieces touching the elections: the Roy Cooper Senate race in North Carolina, an op-ed arguing that Trump's Jones Act waiver in Michigan could cost Republicans the Senate, and a feature on candidates using sports references to connect with voters. The Jones Act piece was the most substantive — an argument that a specific maritime-policy decision in the Great Lakes is alienating union voters in a state where the Senate margin may be razor-thin.
That is a granular, inside-baseball entry point: policy mechanics and regional vote math. It reaches a reader differently than a national horse-race frame would. Eight outlets covered the midterms in this set, tilted center-right — a reminder that which outlets show up on a story is part of the framing before a single word is written, which is why we keep arguing that source count matters more than any single article.
How does media framing affect public perception?
Lined up together, these four stories make the answer concrete. Framing rarely changes the facts. It changes which facts arrive first, which actor the story is "about," and whether an event feels like an earthquake or a footnote. A reader who only saw CNN's Serbia headline and a reader who only saw the Financial Times' both learned that Vučić conceded — but one learned it mattered and one learned it happened.
The defense against that is not finding the one unbiased outlet. It is reading across the spread and noticing the frame itself. A few patterns were worth flagging this week beyond the four stories above. The set logged a left-side blindspot on the neocloud-stock selloff, where Signal/noise noted that left-leaning outlets skipped a pension fund's retreat from AI — "CalPERS pulling back from big tech is exactly the kind of story that outlets covering labor, public finance, and economic inequality typically pursue." And on Israel's upcoming election, only Axios reported that Netanyahu had personally called Trump for a public endorsement in the days before the vote — a detail five other previews left out.
If you want to practice spotting this on your own, our guide to detecting media bias and the breakdown of left-leaning versus right-leaning source ecosystems are the places to start. Or open this week's daily digest and read two outlets on the same story back to back — the frame tends to show itself by the second paragraph.