Fact Check: How a Mislabeled Chart About Bisexual Women Became 'Proof' of Trans Social Contagion

In March 2026, a chart appeared on X claiming to show the percentage of young people identifying as transgender — a steep upward curve that looked alarming by design. Elon Musk amplified it within five minutes. Combined views exceeded 30 million in under 48 hours. The replies filled with variations of "social contagion confirmed." The problem? The chart actually showed the percentage of US women identifying as bisexual, broken down by age group, from CDC data analyzed by psychologist Jean Twenge. The original label had been stripped off and replaced with a false claim about transgender youth.

This is a story about one chart. But the skills you need to catch this kind of manipulation will outlast every news cycle.

What the chart actually showed

The original data came from the CDC's National Survey of Family Growth and related behavioral surveys, spanning 2014 to 2025. Psychologist Jean Twenge — known for her research on generational trends — published an analysis on her Substack examining the rising percentage of young American women identifying as bisexual. The chart plotted this trend by age group, showing a significant increase among women aged 18–25 over the past decade.

The data is real. The trend is real. And it has nothing to do with transgender identification.

Twenge's analysis explored several possible explanations for the bisexual identification trend — including reduced stigma, broader definitions of sexuality, and social influence effects. It was a nuanced, data-grounded discussion. What it was not, in any way, was a chart about transgender youth.

How the mislabeling happened

The mechanics were distressingly simple. Someone screenshotted the chart, cropped out Twenge's original title and axis labels that referenced bisexual women, and added a new caption: a claim that it depicted young people identifying as transgender. The visual shape of the data — a rising curve — did the rest. No source link. No methodology. Just a graph that "looked scientific" paired with a politically charged claim.

By the time Musk reposted it, the false framing was baked in. Millions of people saw a chart with a steep upward trend and a caption about trans youth. Most never questioned it. Why would they? It had a graph. Graphs are data. Data is truth. Except when it isn't.

The Journal, Ireland's fact-checking outlet, was among the first to trace the image back to Twenge's original Substack post and confirm the mislabeling. Their investigation showed a clean chain: original chart about bisexual women → cropped and relabeled → posted as evidence of trans social contagion → amplified to tens of millions.

How to verify any chart you see online

This incident is a case study, but the verification method works on any data visualization you encounter. Here's a four-step process:

1. Find the original source

Every legitimate chart comes from somewhere — a study, a dataset, a report. If the post doesn't link to a source, that's your first red flag. Search for key phrases in the chart's text, or use reverse image search (Google Lens, TinEye) to find where the image first appeared.

2. Check the axes and labels

What exactly is being measured? What are the units? What's the time range? A chart without clearly labeled axes is, at best, incomplete. At worst, it's intentionally vague. In this case, the original Y-axis label referencing bisexual identification had been removed — the single change that made the entire deception possible.

3. Trace the methodology

Who collected this data? How? What was the sample size? CDC surveys follow rigorous methodology. A random chart on social media does not. Understanding where data comes from tells you how much weight it deserves.

4. Check who's sharing it and why

Context matters. A chart shared by a researcher with a link to their methodology is fundamentally different from the same chart stripped of context and shared by an account with a political agenda. The data didn't change. The framing did.

These skills aren't just for catching misinformation about social issues. The same techniques apply to manipulated charts about AI capabilities, industry benchmarks, or technology comparisons. Wherever data visualizations are used to persuade, these verification steps apply.

Why data visualizations are uniquely dangerous when weaponized

Charts occupy a special place in how we process information. They carry an implicit authority that text doesn't. A written claim — "transgender identification is skyrocketing among youth" — invites skepticism. The same claim presented as a line chart with axes and data points feels like evidence. Our brains treat visualized data as more objective, more scientific, more trustworthy than words alone.

This is what makes mislabeled charts so effective as misinformation. They exploit the trust we place in data visualization as a medium. You don't need to fabricate data. You don't need to hack a government database. You just need to take a real chart, remove the label, and add a new one. The visual does the convincing for you.

Research in cognitive science supports this. Studies on the "scientific imprimatur effect" show that people rate claims as more credible when accompanied by graphs — even when the graphs contain no relevant information. The format itself is persuasive.

The liar's dividend

There's a second-order effect that's arguably worse than the misinformation itself. When people learn that charts can be faked or mislabeled, some begin to doubt all data visualizations — including accurate ones. Researchers call this the "liar's dividend": the existence of fakes gives people permission to dismiss real evidence that contradicts their beliefs.

This is the deeper cost. The mislabeled bisexual-women-to-trans-youth chart didn't just mislead 30 million people about one topic. It contributed to an environment where legitimate data — from the CDC, from peer-reviewed studies, from careful researchers like Twenge — gets dismissed with "you can't trust any chart you see online."

The answer isn't to stop trusting data. It's to get better at reading it.

The bottom line

The chart was real. The data was real. The label was a lie. And 30 million people saw the lie before anyone checked. In an information environment where a single screenshot can outrun every fact-checker on the planet, the most important skill isn't knowing the right answer — it's knowing how to ask the right questions before you hit share.

Frequently asked questions

What did the viral chart actually show?

The chart showed the percentage of US women identifying as bisexual by age group, from 2014 to 2025. The data came from CDC surveys and was originally published by psychologist Jean Twenge on her Substack. It had nothing to do with transgender identification. Someone removed the original labels and added a false caption claiming it showed trans youth data.

How can I tell if a chart on social media is real or manipulated?

Use a four-step check: (1) look for the original source — if there's no link, be skeptical; (2) read the axes and labels carefully for specifics; (3) trace who collected the data and how; (4) consider who is sharing it and whether context has been removed. Reverse image search can help you find the original version.

What is the "liar's dividend" in misinformation?

The liar's dividend is a concept where the known existence of fakes and manipulated content gives people a reason to dismiss real evidence they find inconvenient. Once people learn that charts can be mislabeled, some use that knowledge not to become better readers of data, but to reject any data that challenges their existing beliefs.

Sources:

• The Journal (Ireland) — Fact Check: Viral chart does not show transgender identification among young people
• Jean Twenge — Original analysis of CDC data on bisexual identification trends, published on Substack
• CDC National Survey of Family Growth and Youth Risk Behavior Survey data (2014–2025)