Influence across cultures: four questions to ask of any claim

Someone hands you a briefing on how a country negotiates, a slide on how a region treats deadlines, or a summary of how a group responds to directness. Whether the claim is true is the wrong first question. Ask whether it could tell you anything useful about the person you are meeting on Thursday.

Usually it cannot, and the reasons are structural rather than a matter of the work being done badly. Four questions test any such claim using nothing but the claim itself, and a short checklist then decides what a briefing may keep.

What to take away

Who was measured, and when

The first question is who was actually measured. Very often it is people who were easy to reach: employees of large multinational firms, students at one institution, members of a professional association. That is convenience sampling, and it is not a scandal.

It does mean the result describes a particular slice rather than a country. A finding drawn from corporate managers tells you about corporate managers, so a briefing that names no sample has given you nothing to apply to a supplier, a client or a new colleague.

The same questions apply to country score tables such as Hofstede's cultural dimensions or the GLOBE project: ask what was measured, when, and in which sample before a line from them goes into a briefing.

The second question is when. Working norms change with generations, with technology, and with who a country's employers are. A description of how business was conducted somewhere twenty years ago may be accurate about then and useless about now. Nothing in the claim itself will tell you which.

Compared with what, and would it travel

A statement that a group is direct, or hierarchical, or slow to commit, is meaningless without a comparison. Compared with whom, on what measure, and with how much overlap between the two distributions? Overlap is the number that matters and the one almost never quoted.

The last question is whether the pattern would survive being moved. A pattern found in one setting may or may not appear in another, which is the question of external validity. A finding about a laboratory task, or about one industry, is not automatically a finding about a supplier negotiation in a different sector.

What each kind of claim can support

A claim can be true and still unusable. Each row pairs a common statement with what it supports and what it does not, and the right-hand column is where practical use collapses.

Every row can be true and none of the rows tells you what to do on Thursday, because a fact about a distribution is not a fact about a draw from it.

A checklist before a briefing circulates

Run every line of a briefing or training slide through five checks before it goes to the team. A line that fails one is not wrong, only unfinished, and the verdict should travel with the line so the next reader sees why it changed.

Worked through, a slide that says "this country prefers indirect communication" fails four checks at once: no sample, no date, no comparison group and no setting. It survives only as a watchlist line, such as "norms differ on how disagreement is voiced, so watch for it".

What survives the audit

Two uses of a general claim are honest. The first is a watchlist: groups differ on whether a meeting decides or ratifies, so you know to look. That tells you nothing about which one this is, and the briefing should say so.

The second is institutional, not behavioral: how contracts work, what a regulator requires, what the working week is. Those are facts about a place, and they can be looked up. They belong in a briefing in a way that a prediction about a person does not.

What does not survive is the predictive use. Once you carry an expectation about a person into a room, you will find it confirmed, because that is what expectations do to observation, and you will stop collecting the evidence that would have corrected it.

The method that replaces prediction is set out in influence across cultures, and the four senses of culture help you decide which one a claim is about. For what changes between workplaces and what does not, see working across cultures compared.

Much of what gets attributed to culture is a response to seniority, described in leadership influence.

Common questions

This sounds like a rejection of the whole field.

It is a rejection of one use of it. Work that documents how norms vary is worth having, and the same work almost always says explicitly that it describes populations rather than individuals. The misuse happens downstream, in the summaries and the one-page briefings.

My employer's training uses country scores. What do I do with that?

Take the list of dimensions as a checklist of things to observe and discard the country values. The dimensions are a useful vocabulary for what varies. The scores are the part that cannot tell you about the person you are meeting.

How large does a difference have to be before it is useful?

Large enough that the distributions barely overlap, which is rare for anything measured about human behavior. The overlap, not the gap, is what determines whether an average can predict an individual.

What should I do with a briefing that gives no sample?

Treat it as an opinion with a number attached. Ask the author who was measured, when, and against what. For errors that survive experience, see the mistakes experienced operators still make.