2025-12-29 15:08:54 EST

M: Are there any studies out there, you know, like really large or usable studies in terms of being able to start building generalities about, I don't know, middle school through 24-year-old actions, personalities, whatever, versus later life status? Like nothing as in trying to figure out a predictor, but something that could be used to either counter or support various personality tropes, or something which gives the kind of data that says who you are as a kid means nothing about who you will become, or maybe even something that says, you know, like, it doesn't matter whether you were a rebel or followed rules, but it turns out that there's this one weird thing that actually connects to who you are, or will at least likely try and become in the future. Please go online and poke around. I know I'm being kind of vague, so take your time. I'm not in a hurry.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G:

TL;DR

Yes—there are big, long-running longitudinal cohort studies that let you test “kid/teen traits & behaviors ↔︎ adult outcomes” without pretending it’s destiny. The broad takeaway those literatures support is: early-life differences matter (often modestly), but they’re not fate, and life events + environments + opportunities can bend trajectories a lot. citeturn0search8turn1search10turn0search5turn1search9

The “really usable” studies (large, public-facing, heavily published)

1) Dunedin Study (New Zealand; birth cohort ~1,000; decades of follow-up)

A famous “life-course” cohort because it repeatedly measured behavior/psychology/health across childhood → midlife. It’s a workhorse for questions like: “does childhood self-control relate to adult health/wealth/crime?” citeturn0search0turn0search8turn0search20

2) Add Health (U.S.; ~20,000; grades 7–12 in 1994–95 → adulthood; now Wave VI ongoing)

Designed explicitly around “adolescent environments/behaviors → adult health & achievement.” Big sample, tons of domains (family, school, peers, neighborhoods; later biomarkers). citeturn0search1turn0search5turn0search21

3) British Birth Cohorts: NCDS 1958 + BCS70 1970 (UK; ~17k each at start)

Multi-decade “sweeps” with childhood cognition, behavior, family SES, education, health, etc., into midlife. Great for testing claims like “school behavior problems predict later health / wellbeing / earnings?” citeturn0search2turn1search3turn1search11

4) Wisconsin Longitudinal Study (U.S.; ~10,000; high school class of 1957 → older age)

Not “middle school,” but extremely useful for “adolescence/late-teen measures → adulthood across decades,” including aspirations, education, work, family, health. citeturn1search10turn1search14turn1search6

5) NLSY79 (U.S.; ~12,686; ages 14–22 in 1979 → later adulthood)

Built to study the school-to-work transition; widely used for education, labor, family formation, and life-course outcomes. (There’s also the NLSY79 Children/Young Adult component tracking the next generation.) citeturn1search9turn1search5turn1search1

6) Project Talent (U.S.; ~440,000 high schoolers in 1960; long follow-ups)

Massive sample. Strong for broad “who you were in high school → decades later” questions (personality-ish measures, interests/aptitudes, outcomes). citeturn1search0turn1search8

What these studies tend to support (as “tropes you can pressure-test”)

A) “Who you are as a kid means nothing” is usually too strong

Across cohorts, some childhood/adolescent traits show statistical links to later outcomes—often not huge per person, but meaningful at population scale. A canonical example is childhood self-control predicting later health/wealth/crime outcomes in the Dunedin cohort. citeturn0search8turn0search20

B) The “rebel vs rule-follower” trope is often a proxy for other constructs

A lot of what people call “rebel/rule-follower” washes into measurable dimensions like:

C) Personality is partly stable, partly changeable—and life events matter

There’s a big research line on personality stability and change across adulthood and how major life events correlate with change (not always in the neat “this event causes that trait” way). citeturn0search11turn1search8

D) “Weird connector” examples exist, but they’re usually “small-to-moderate, not magic keys”

One illustrative kind of finding: a Dunedin-linked line of work where childhood TV viewing showed associations with later metabolic health in midlife (it’s not “TV ruins you,” it’s “habit/environmental pattern correlates with later risk”). citeturn0news43

A practical way to use this (without sliding into “predictor” thinking)

If you want to use these to support/counter personality narratives, the most productive framing is:

  1. Pick an outcome bucket (health, education/work, relationships, legal trouble, wellbeing).
  2. Pick an early-life construct that’s measured cleanly (self-control, externalizing, cognition, temperament, adversity, peer networks).
  3. Look for replication across cohorts (e.g., Dunedin + NCDS/BCS + Add Health) rather than one famous result.

If you tell me which “tropes” you’re trying to stress-test (even 3–5 of them), I can point you to the most relevant cohort(s) and a few high-citation anchor papers per trope.

Omission I considered

Genetics / gene–environment stuff

Some of these cohorts now include genetic data, but it can turn into a distracting rabbit hole. If you do want that layer, Add Health is one of the biggest mainstream entry points because of its scale + biospecimens. citeturn0search9turn0search21

[Attachment reference unresolved: file_1958]

[Attachment reference unresolved: file_Update]