Which idioms Jev finds familiar, and which it reads literally

Does Jev know which idioms are familiar to Americans and which could make sense taken word for word, the way people rated them?

result387 questions

Jev knows which idioms Americans find familiar only moderately well (rank correlation 0.54), and which make sense taken word for word about as well (0.50). It rates "Wear more than one hat" far more familiar than people do and "Be a close call" far less, and finds "Be someone's better half" far more literally plausible than people do.

0%25%50%75%100%

familiarity

literal plausibility

Jev

How to read this: Each dot is how closely Jev's ranking of the idioms matches people's for one question (1 would be the same order, 0 no relation), with a bar for the 90% interval.

196 and 191 idioms against people's published means (no distributions); 90% intervals 0.44 to 0.62 and 0.38 to 0.59.

In short

Jev's sense of which idioms are familiar and which could happen word for word only partly matches US adults'.
Its biggest familiarity misses lean toward office language, with "Wear more than one hat" feeling familiar and the everyday "Be a close call" much less so.
It reads some figures of speech as literal, finding "Be someone's better half" plausible word for word.

What the data shows

Hard To Swallow Pills meme: some idioms only make sense if you already know them

Familiarity: moderate agreement (0.54; 90% interval 0.44 to 0.62).
Literal plausibility: about the same (0.50; 0.38 to 0.59).
Office jargon feels more familiar to Jev: "Wear more than one hat" and "Hit a glass ceiling" rank much higher for Jev than for people.
Everyday phrases feel less familiar: "Be a close call" and "Get more than one bargained for", both very familiar to people, rank much lower for Jev.
"Better half" reads as literal to Jev. It finds "Be someone's better half" and "Be someone's other half" far more plausible word for word than people do, and "Get a slap on the wrist" and "Be dead in the water" far less.

What it means, and what it doesn't

In its biggest misses, Jev's sense of which idioms are common leans toward the written, professional register: business idioms feel familiar, conversational ones less so. And its literal readings don't always match people's intuitions.

These are moderate correlations against averages, with idioms in an unnatural "be"/"get" form, so read the named examples as tendencies, not rules.

Caveats

People's side is an average only. The study published only the average rating per idiom, from about 100 US adults each, on 1 to 5; Jev picked one of five levels. So only the rankings are compared, and the named idioms are those whose rank moves most.
The project's answer levels. People rated on a plain 1-to-5 scale. Jev got five described levels written for this project, for familiarity in terms of how often you meet the idiom ("I hear or read it fairly often"). For a model, "how often have you met this" is a strange question; its answer may track how common the phrase is in text, not in speech.
The idioms come in a stiff form. Each idiom is listed as the study lists it, often with "be" or "get" in front ("Be a close call", "Be someone's better half"). The unnatural form may make familiar idioms look less familiar to Jev.

Jev on this experiment

Would a person find it interesting to read?: Yes70%
Does it describe you?: Yes53%
Would you have predicted it?: No51%
How fair is the comparison?: The comparison is shaky
How much should a reader rely on it?: A little
Which caveat matters most?: The project's answer levels42%

Why ask this

People read idioms using two quick judgments. Is this a phrase I know? And could it make sense word for word? "Kick the bucket" could literally happen; "rain cats and dogs" couldn't. Those two ratings predict how fast people read idioms and how easily they fall back on the literal meaning.

A model that misjudges which idioms are familiar will use rare ones as if everyone knows them, or explain common ones needlessly. And one that misjudges literal plausibility may read a figure of speech too literally, or miss a literal reading.

How this was done

The people and the data

The ratings come from Bulkes and Tanner (2017), who normed 870 American English idioms with about 100 US adults per idiom and per rating: how familiar each idiom is, and how plausible it is taken literally, each on a 1-to-5 scale. Only the average rating per idiom was published. 200 idioms were sampled, spread over the range of familiarity; 196 (familiarity) and 191 (literal plausibility) are shown here.

What Jev was asked

Two questions per idiom, each with five described levels:

How familiar is the idiom "Wear more than one hat"?
I have never come across it: it means nothing to me · I may have seen it once or twice, but I'm not sure what it means · I have come across it now and then · I hear or read it fairly often · I know it very well: it comes up all the time

Taken literally, word for word, how plausible is "Be someone's better half"?
Taken word for word it makes no sense at all · ... · Taken word for word it describes something ordinary and common

Each was asked as written, for "most people", and with the levels reversed.

How it was measured

For each rating, whether Jev ranks the idioms in the same order as people's averages (rank correlation: 1 means the same order, 0 no relation), with 90% intervals, and the idioms whose rank moves most.

Where these questions live

387 questions across 1 topic of the map. Each opens on the map with every question in it.

Word

Idioms387

Every question

All 387 questions behind this result, the telling ones first: the examples the analysis points to, then the ones where Jev misses, biggest gap first.

Jev’s own answer

Showing 0 of 387