Informational influence research explained with examples

Somebody tells you that people copy others because they assume the crowd knows something, and gives you a figure for how often. Before using the figure, work out what would have had to be true for anyone to know it.

What to take away

• The direction is well supported and the magnitude is disputed. Numbers that circulate confidently usually come from one narrow setting.
• The most useful idea in this literature is structural rather than numerical: information can look independent and not be.
• Where a claim cannot say who was asked and what they could see, treat it as a lead rather than a finding.

What is not seriously in doubt

People do use other people as evidence, and they are right to. Almost everything you know arrived through somebody else, and the alternative, checking everything personally, is not available to anyone.

It is also clear that this can go wrong in a specific structural way. When each person can see what others did but not why they did it, the reasons stop traveling and only the behavior does. A large agreement can then form out of a small amount of original evidence, and it will feel to every participant as though many people independently reached the same conclusion.

That structural point does not depend on any particular measurement. It follows from the setup, which is why it is the part of this literature worth carrying around.

What is genuinely disputed

• How strongly people weight others relative to their own information. Estimates vary widely between settings and the variation is not noise.
• Whether laboratory arrangements transfer to ordinary life, where the stakes, the audience and the option to walk away are all different.
• How much of the copying is people treating the crowd as evidence and how much is people not wanting to be the odd one out. Those are separate channels and they are difficult to separate in practice, which is why the second one has its own account under normative influence.
• How durable any of it is. A shift measured minutes after an exchange and a shift that persists for a week are different results, and the first is much easier to produce.

Anyone giving you a single confident number across all of that is compressing a live argument into a decoration.

The questions that make a claim usable

Question	Why it decides the weight
Could participants see the reasons, or only the choices?	This determines whether independence was ever possible
Did people know they would be observed?	Being watched turns an evidence question into a social one
What was the cost of being wrong?	Free choices and costly ones behave differently
Was the finding measured once, or later as well?	Immediate shifts often do not persist
Has anyone reproduced it somewhere else?	A single striking result is a lead

The last row carries the most weight. Across behavioral fields, a substantial share of memorable results have proved harder to reproduce than expected, and what that means for any individual claim is argued over rather than settled; the Stanford Encyclopedia entry on reproducibility of scientific results lays out the positions. The practical rule for a reader is modest: one memorable result is a reason to look further, never a quantity to plan around.

The trap that matters more than any number

Circular reporting. Three sources agree, and all three are repeating one original. Nothing is fabricated at any step, and the agreement carries no additional evidence at all.

This is the single most useful thing to take from this area, because it can be checked. Follow each source back one step. If they converge on one origin, you have one source with three voices, and the confidence you drew from the agreement was borrowed from a repetition.

The same shape appears in expert consensus, in press coverage, and in your own social circle. It is not a rare pathology; it is the default outcome whenever transmission is cheap and attribution is not. What you are entitled to believe on the strength of other people is the standing question in social epistemology, and this is its most practical corner.

What the evidence cannot carry

It cannot tell you whether the crowd in front of you is right. Structural facts about how agreements form say nothing about whether this particular agreement tracks reality.

It cannot settle the ethical question, which is prior and separate. A technique that reliably moves people is not thereby permitted, and the standard this site uses does not depend on effectiveness: influence that survives the other person learning how it worked is legitimate. That test is set out under social influence basics and does not move when the evidence does.

And it cannot be run backward on yourself. Knowing that people copy others tells you nothing reliable about whether you copied on a particular occasion, which is why the workable habit is a written record before the conversation rather than an introspection after it. The mechanism itself is set out under informational influence, the version that runs on published counts under social proof, and the deliberate exploitation of all of it under manipulation resistance.

Common questions

Why does this page give no figures?

Because a figure without its setting is not information here, and quoting contested numbers would trade your accuracy for a feeling of precision.

Does disputed evidence mean the effect is not real?

No. Direction and magnitude are separate claims. The direction is supported by ordinary observation as well as by research; the size is what is argued about.

How do I check for circular reporting quickly?

Take each source back one step and see where they land. Two minutes usually settles it, and the answer changes how much the agreement is worth more than any further reading would.

What should I look for in a source that summarizes this area?

Whether it reports the setting and the uncertainty alongside the result. A summary that hands you a clean number with no conditions has removed the part you needed.