Closeness-Communication Bias

What it is

Most people hold that they communicate better with close friends than with strangers. The authors do not dispute that this is often true. Their proposal is narrower: closeness can also raise a speaker's confidence faster than it raises the listener's understanding, so the speaker misjudges how much has got across.

The explanation they offer rests on perspective-taking. With a stranger, a person assumes the other's view probably differs and monitors it. With a friend, a person assumes similarity, relaxes that monitoring and leans on their own perspective as the default. The authors place this in a two-step picture of social judgement: an egocentric starting point, followed by an effortful correction that is weaker for close others. They add that perceived overlap with a friend may grow faster than actual overlap, and that a person may carry an assumption of shared understanding from one area of a friendship into another where it does not hold.

The idea sits next to the curse of knowledge. Both describe a speaker or listener leaning on their own information when judging someone else's. The curse of knowledge concerns what a person knows. The closeness-communication bias concerns who the other person is.

In effect

The first experiment tested listeners. University of Chicago undergraduates acted as directors, who gave instructions, and addressees, who followed them by moving objects in a grid while their eye movements were recorded. Some grids held a hidden object that only the addressee could see and that could be confused with the intended one. Each director took part twice, once with a friend and once with a stranger. When the hidden object was relevant, addressees paired with a friend were slowed more, took longer to correct an early look at it, and reached for it more often: on 24 per cent of trials against 15 per cent with a stranger. Both groups glanced at the hidden object about equally often, so the difference lay in correcting the error and not in making it.

The eye-tracking measure matters because it separates two things a simple error count would merge: whether the listener is drawn to the hidden object at all, and whether they notice and correct it in time. The friend and stranger conditions looked alike on the first and differed on the second.

The second and third experiments tested speakers. Each speaker worked through 20 ambiguous phrases, every phrase carrying four possible meanings, with three listeners across four rounds, so chance performance was one in four. In the first, 60 Williams College students in friend pairs tried to convey one of several meanings of ambiguous phrases to listeners, and predicted how many listeners would understand them. They expected friends to understand them far better than strangers, while friends were only marginally more accurate. In the second, 24 married couples did the same task. Spouses and strangers did not differ reliably in accuracy, yet speakers were more confident that their spouse had understood, and confidence did not track accuracy. Listeners in both studies scored above chance.

For work, the practical reading is the authors' own logic carried to a new setting. A manager briefing a long-standing colleague, or an expert handing a task to a trusted successor, is in the position of the friend and spouse speakers. The experiments did not test colleagues or workplace documents, so this remains a prediction from the finding and not a result of it.

What it does not say

The authors state that the findings do not show friendship hinders communication. Closeness can make people more comfortable asking for clarification, and shared perspective can make an egocentric default efficient when the overlap is real. The bias is most likely where overlap is overestimated, or where a shared frame from one domain is assumed to carry over to another.

The samples were university students and married couples, tested on directed object movement and ambiguous phrases. The paper does not claim that the pattern holds for every kind of message or every kind of relationship, and it leaves open whether prompting people to think about differences between themselves and a friend would restore the correction that closeness weakens.

Who developed it

Kenneth Savitsky, then at Williams College, was first author. His co-authors Boaz Keysar, Nicholas Epley and Travis Carter were at the University of Chicago, and Ashley Swanson at the Massachusetts Institute of Technology. The paper appeared in volume 47 of the Journal of Experimental Social Psychology.

---