What is a cross-sectional study?

Where a longitudinal study follows one subject across time, a cross-sectional study does the opposite, it snapshots many subjects at once. Checking the citation rate of fifty competing brands this week, or checking robots.txt files across tens of thousands of domains on a single crawl, is a cross-sectional study. It tells you the shape of the landscape at that moment.

The limit is built into the design: a cross-sectional study can't tell you whether any given subject's position is rising, falling, or stable, because it only has one point per subject. It's easy to mistake a snapshot ranking for a trend, treating "we're ahead of competitor X this week" as a durable fact rather than a single reading that could look different next week.

The two designs answer different questions and neither substitutes for the other. A cross-sectional study is how you'd learn that 16.3% of readable domains block at least one AI crawler; it can't tell you whether that share is growing or shrinking without a second crawl months later to compare against.