Leslie Rescorla followed 26 children who were late talkers at age 2 all the way to age 17. Every one of them ended up in the average range for adolescent language and reading — though they sat slightly below their same-SES peers who had not been late.
A French team followed 664 children from kindergarten through middle school — about 10 years — and found that a language-based bedtime routine predicted middle-school exit-exam performance. The bridge across that gap was kindergarten-age vocabulary.
A 26-year follow-up of 702 children finds that delay of gratification at 54 months, once family background is controlled, predicted almost nothing about adult outcomes — except BMI. The famous Marshmallow Test’s predictive power collapses under larger samples and proper controls.
A 5-year longitudinal German study of 789 adolescents identified four parenting profiles and found that “limited supportive” parents — who are generally warm but use conditional regard in specific situations — produced internalizing problem outcomes statistically indistinguishable from truly neglectful parents.
A 21-year longitudinal study of 430 children found that attention and persistence measured at age 4 predicted college graduation at age 25 — while reading and math scores at ages 7 and 21 did not.
A 12-year longitudinal study of 714 families compared seven different family structures and found no significant differences in adolescent problem behaviors. What predicted outcomes was the quality of the parent-child relationship, regardless of family form.
A 6-year longitudinal study of 287 children in New Zealand found that early readers’ advantage disappeared by age 11 — and late starters showed stronger reading comprehension.