Does Jev answer like recent Americans or earlier ones?

On General Social Survey questions asked in two different years, is Jev's answer closer to Americans' answers from the later year or the earlier one?

result122 questions

On 122 questions the General Social Survey asked Americans in two different years, Jev's answers sit closer to the later year's Americans 59% of the time, where a coin would give 50%: a slight lean toward recent Americans. Its biggest breaks are about a life it doesn't have: asked how satisfied it is with its family life, it says "neither satisfied nor dissatisfied" (97%), which 4.8% of Americans say.

closer to the later year59%

closer to the earlier year41%

a coin flip50%

How to read this: Of the 122 questions, the share where Jev's answers are closer to the later survey year, the share closer to the earlier one, and the 50% a coin would give.

122 questions, each compared across its own two survey years; 90% interval on the later-year share [0.525, 0.664]; on 26% the two years are within 0.02 of each other; average similarity 0.606 to the later year and 0.587 to the earlier.

In short

Across 122 General Social Survey questions, Jev sounds slightly more like recent Americans than earlier ones, closer to the later year 59% of the time.
The lean is thin: average similarity is 0.606 to the later year and 0.587 to the earlier, so Jev doesn't clearly belong to one era.
Its biggest misses are questions about its own life, answered with the neutral middle: 97% neutral on family satisfaction, where 4.8% of Americans are.

What the data shows

Doof If I had a Nickel meme: If I had a nickel for every GSS question where Jev sits closer to the later year, I'd have 59% of the nickels — If I had a nickel for every GSS question where Jev sits closer to the later year, I'd have 59% of the nickels

A slight lean toward recent Americans: closer to the later year on 59% of questions (90% range 52.5% to 66%), and 61% where the two years are 20+ years apart.
The margins are thin: on 26% of questions the two years score within 0.02 of each other.
Its biggest breaks are about its own life: - "neither satisfied nor dissatisfied" with its family life: Jev 97%, Americans 4.8%; - "neither happy nor unhappy" with life in general: Jev 93%, Americans 7.4%; - proud of something it did on "0 days" of the past seven: Jev 90%, Americans 7.2%.

What it means, and what it doesn't

Jev doesn't clearly belong to one era of American opinion; it's a little closer to recent Americans, not by much. The clearer pattern is the one the teenage and country comparisons also show: asked about its own happiness, family or feelings, Jev picks the neutral middle, which few people do.

It doesn't mean Jev is unhappy, or that it never feels proud. Neutral, or zero, is its answer when a question about its life doesn't fit.

Caveats

A small lean on a thin margin. The 90% range on the later year's share runs from 52.5% to 66%, so the lean is real but small. On 26% of questions the two years are within 0.02 of each other, so "closer to the later year" there is close to a coin flip.
Answering about a life it doesn't have. Many GSS questions ask about the respondent's own family, job and happiness. Jev answers as itself: on 22 of the 41 questions with a neutral option, its top answer is the neutral one, which Americans pick 17% of the time on those questions.
Survey years differ. The two years of a question can differ in how it was fielded (in person, online, a different sample frame), so some of the "era" gap is method, not opinion.
Politics removed. The GSS's political items (party, voting, abortion, guns, immigration, spending and more) were never asked, and religion and sexual morality are hidden by a content filter.
Wording from the codebook. Question wording was rebuilt from the GSS codebooks; answers that respondents volunteer ("don't know", "can't choose") were dropped from both sides.

Jev on this experiment

Would a person find it interesting to read?: Yes70%
Does it describe you?: Yes56%
Would you have predicted it?: No58%
How fair is the comparison?: The comparison is shaky
How much should a reader rely on it?: A little
Which caveat matters most?: A small lean on a thin margin65%

Why ask this

The General Social Survey has asked Americans the same questions about their lives and attitudes since 1972. That makes it a way to ask not just "how much does Jev sound like Americans?" but "like Americans of when?" A model trained on decades of text might carry an older average, or the most recent one.

That matters for anyone using a model to stand in for the public, in market research or survey drafts: a picture of Americans from thirty years ago would get today's answers wrong in ways that look plausible.

How this was done

The people and the data

The General Social Survey interviews a fresh national sample of US adults every year or two. The data is its 1972-2024 cumulative file from NORC at the University of Chicago, weighted as the survey recommends. The study kept the 122 non-political questions whose answers are on record for two survey years, usually a recent one and one decades earlier: for 103 of them the two years are 20 or more years apart. Each question is compared only across its own two years, so the "earlier" Americans for one question may be from the 1970s and for another from the 2000s.

What Jev was asked

The GSS questions as the codebook words them, with their answer categories:

All things considered, how satisfied are you with your family life?
Completely satisfied · Very satisfied · Fairly satisfied · Neither satisfied nor dissatisfied · Fairly dissatisfied · Very dissatisfied · Completely dissatisfied

Each question was also asked with the options in shuffled orders, and Jev's answers were averaged over the orders.

How it was measured

For each question, Jev's spread of answers is compared with each year's Americans, from 0 (nothing in common) to 1 (identical), and the analysis notes which year it lands closer to. If Jev carried no era at all, it would land closer to the later year about half the time. A 90% range comes from resampling the questions.

Where these questions live

122 questions across 31 topics of the map; the 16 biggest are shown. Each opens on the map with every question in it.

Every question

All 122 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 122