Crossed answers
The move that looks like rigour
Of all the ways to answer badly, citing evidence is the one that raises your standing while lowering the quality of the discussion.
Claims this part is about, and what each was answered with
A company is a place, not a login
Value≠A studyYoung people are owed the experience of working near others
Value≠A studyPeople are less likely to leave if they feel connected to a place
Empirical≠A principleHybrid teams make worse decisions than co-located ones
Empirical≠A principlefour claims · two empirical and two value · four were answered with the wrong kind of answer
Four of the twelve crossed answers on this record.
There is an uncomfortable fact about crossed answers: the commonest one is also the one that makes the person doing it look best.
Producing a study at a value claim is the move that reads as the most rigorous contribution in the room. It is also inert, and the combination is why it is so hard to stop.
What it looks like from the outside
Somebody has read something. They are bringing evidence rather than opinion. They are raising the standard of a discussion that has been full of assertion.
Every one of those perceptions is reasonable and none of them depends on whether the evidence bears on the claim — which nobody in the room is checking, because checking it requires sorting the claim, which nobody is doing.
So the reputational return on citing evidence is roughly the same whether or not the citation is apt.
Who this catches
Disproportionately, the most conscientious people in the discussion.
Somebody who does not read research does not make this mistake. The person who does it is the one who went away, found the literature, read it and came back — and the better they are at that, the more authority the inapt citation carries.
On this record the person who cited most often was also the only participant who had read any of the papers properly, could summarise them accurately, and knew their limitations. They were crossing claims all the same, because reading a paper well and checking that it addresses the claim in front of you are different operations.
The test, before you cite
Two sentences, written or thought, in this order:
This study shows __ about ____.
The claim I am answering is about ______.
If the two blanks at the end are different things, the citation does not work. It took about fifteen seconds to apply to each of the four most-used citations on this record, and three of them failed.
What to do with a study that fails the test
Not nothing. The study is usually about something real and worth saying, and there is a form that keeps it:
This is not about the thing you said, but it is interesting on a related point — it finds X at the firm level, and our question is about teams. Worth knowing, not an answer to you.
That sentence does three things at once. It keeps the contribution, it declines the inference explicitly, and it models the sorting without lecturing anybody about it.
The harder version
There is a further step, and it is the thing that most improves these discussions when anybody does it.
Produce the study that goes against you.
There is a paper that cuts the other way on this and I think it is the better of the two. Here it is.
Nothing else so reliably changes the temperature of a room, and the reason is not generosity. It demonstrates that you are treating the empirical part as genuinely open, which is the only thing that makes your other citations worth attending to.
The status problem
Underneath this is a problem about how rooms assign credit, and it is not solvable by individuals.
Evidence carries status. Values do not. So a room systematically rewards the empirical register regardless of whether the claim in front of it is empirical, and a person who states a value plainly is heard as having nothing stronger to offer.
That is a bad incentive and it produces exactly the pattern on this record: ten value claims, eight of them answered with studies, and several of them dressed as findings in the first place.
What a chair can do about it
The one intervention that works is to credit a well-stated value out loud.
That is useful — you have said what you actually think rather than dressing it as a finding, and it is clearer for it.
Said once, by somebody with standing, it changes what the room rewards for the rest of the session. It is also entirely sincere: a plainly stated value is more useful than a borrowed statistic, and saying so is just accurate.
On this record nobody said anything of the sort, and the empirical register kept its monopoly on seriousness for eleven weeks.
The quiet version of the same problem
There is a less visible form that does more damage over time: not citing a study, but referring to one.
"The evidence on this is pretty clear" does the work of a citation without being one. It cannot be checked, it cannot be found to measure the wrong thing, and it carries nearly the same authority in a room.
On this record the phrase or something like it was used eleven times by five different people, and on no occasion did anybody ask which evidence.
The question that disposes of it is mild and nobody ever asks it:
Which one? I would like to read it.
What to do if you have used it
Most people in this argument have. The repair is to go and find the thing you were referring to, read it, and either cite it properly or stop saying it.
Often there is no specific study — the impression came from a summary of a summary. Finding that out about one of your own claims is the single most useful hour available in this subject.