Other people's behaviour acts as a cue for what to do, buy, or believe.
More quotesAsked privately, 12 per cent selected “subversive activities” as the country’s most important problem. After seeing an apparently unanimous group select it, 48 per cent did the same. The opinion did not gradually mature through public debate. It quadrupled when agreement was placed in view.
That is useful evidence about consensus and bloody inconvenient evidence about surveys, meetings and “most popular” labels. Agreement can record several independent judgements. It can also record one judgement copied several times. The final tally cannot tell you which history produced it.
Consider five estimates spoken aloud. The first is 70. The second person had written nothing down and says 65. The third hears a cluster and chooses 68. By the fifth answer, one uncertain opening has acquired the appearance of five observations. Everyone has behaved reasonably. The record is still misleading.
Private answers before discussion preserve something precious: disagreement before it becomes socially expensive. A range from 35 to 80 is not an embarrassing failure to align. It is a map of hidden assumptions worth examining.
Marketers should be equally suspicious of their own popularity claims. “Most chosen” may describe thousands of separate preferences. It may describe a small early lead that later customers were repeatedly shown. The label works in both cases; its origin disappears.
The experiment reports 48 per cent. It cannot divide that figure into conviction, inference and politeness. Those columns were never collected, so the neat percentage cannot resolve the motive.