Passage Summary: Researchers found that if you don't tell witnesses that the bad guy might not be in the lineup, they are much more likely to pick the wrong person.

Reasoning: Test subjects shown a lineup without the actual culprit were significantly more likely to pick a random person if they weren't warned that the culprit might be absent.

Analysis: The data provided shows a massive drop in false identifications—from 78 percent down to 38 percent—simply by adding a disclaimer. This suggests that the instructions given to witnesses have a profound impact on the accuracy of their choices. We should look for an answer that highlights how witness behavior is influenced by the expectations set by the person conducting the lineup. The evidence strongly implies that witnesses often feel pressured to pick someone unless told otherwise.