I’ve had search summaries cite articles that repeat a statistic without linking to the report or dataset it came from. When I ask Perplexity about a numerical claim, can it follow that trail back to the original material, or does it usually stop at a secondary source?

I’m particularly interested in cases where the source is a public report, a spreadsheet, or a research dataset rather than a news story. How do you check original dataset provenance when the citations look plausible but don’t clearly point to the underlying evidence?