Growth mindset: science

Research on growth mindset runs from Mueller & Dweck (1998) through later critiques. Mechanisms are interacting systems.

Evidence percentages below are relative judgements for this atlas, not fabricated precise statistics. Named studies keep their published bands.

Findings

1998 · Mueller & Dweck — Intelligence praise can reduce challenge-seeking Famous, neat, and later a replication target.
2007 · Blackwell, Trzesniewski, Dweck — Incremental lesson, maths grades A widely cited school study; size and generality later questioned.
2006 · Carol Dweck — Public vocabulary The book created a shared language for teachers — a cultural fact even when effects are small.
2018 · Sisk, Burgoyne, Macnamara et al. — Small average effects Mindset–achievement links and interventions are modest in meta-analysis; some subgroups larger.
2018 · Li & Bates — Replication failures Their studies did not recover the original praise pattern. Methods debate followed.
2019 · Yeager et al. — Small average, larger for some A nationally scaled US experiment: overall small, more for lower achievers in supportive schools.

Interpretation of effort — If effort means “I am dumb”, students hide work. If effort means “I am learning”, they can stay visible.
Goal choice — Learning goals versus looking-smart goals change which tasks get chosen after a success.
Climate lock — A growth message in a school that punishes the first B is a contradiction students can hear.
Not a skill transplant — Believing maths can grow is not a substitute for a well-designed next problem (Ericsson’s hour still has to happen).

Original praise labs

Influential; mixed later replications.

Sisk 2018 average effects

Best sizing of the literature as a whole: small.

Yeager-scale trials

Better powered; still modest averages.

Teacher-language hygiene

Low-cost, theoretically clean: do not treat struggle as a verdict.

Poster-only programmes

Weak as a standalone intervention.

Effect size versus cultural value — A small d can still be worth a sentence of teacher talk. It cannot carry a district’s entire equity plan.
What failed to replicate — Praise experiments, school grade effects, or only some moderators? The camps disagree on the wreckage map.
Mindset and inequality — Critics say the message can become “try harder” without changing load, tracking or tutoring access.

Growth mindset: size the claim to the study. Interaction, not a 50/50 slogan, is the default model.