Psych Reference

---
name: psych-reference
description: "Use when checking what is live (where replication reform actually stands, and the social-media-and-adolescent-mental-health dispute presented as a genuine dispute rather than resolved, verified August 2026), correcting a behavioural-science misconception, finding the canon, or needing a picker and the red flags to check in any behavioural-science claim. Companion to the other psychology-sociology-cultural-sciences skills."
---

# Psychology, Sociology and Cultural Sciences: What's Live, Misconceptions, and Canon

> **Part 5 of 5** of the *Psychology, Sociology and Cultural Sciences* reference (plugin `psychology-sociology-cultural-sciences`), covering §25–§29. Sibling skills: `psych-epistemic-situation-and-replication-reform` (§0–§3), `psych-perception-memory-learning-cognition-and-emotion` (§4–§8), `psych-development-personality-social-clinical-and-neurodiversity` (§9–§15), `psych-sociology-institutions-culture-and-the-weird-problem` (§16–§24). Section numbers are shared across the set; a reference written as §N → `skill` points into that sibling skill.
>
> **Currency:** The core findings are decades stable. Two areas are live. See §25 below for where replication reform actually stands, and the social-media-and-adolescent-mental-health dispute presented as the dispute it is.

> **⚠️ Read §1 → `psych-epistemic-situation-and-replication-reform` first. It is not preamble — it is the single most useful thing in this
> document.**
>
> **In most fields you can learn the findings and treat the epistemics as background.**
> ⚠️ **Here that gets you a head full of confident claims that are false.** **Between
> roughly 2011 and 2020 these fields discovered that a large fraction of their published
> literature did not replicate**, and ⚠️ **much of the popular material about psychology
> was written from the pre-crisis literature and is still in circulation — in
> bestsellers, TED talks, management training and journalism.**
>
> **⚠️ GOTCHA** boxes mark claims that are widely repeated and either false, much weaker
> than advertised, or genuinely contested.
>
> **The three ideas that organize everything below:**
> 1. **⚠️ Knowing WHICH findings survived is more valuable than knowing more findings**
>    (§1–§3 → `psych-epistemic-situation-and-replication-reform`).
> 2. **⚠️ The robust results cluster in a recognizable place**: **perception,
>    psychophysics, memory, learning, individual differences, and large-sample
>    demography.** **The fragile ones cluster in social priming, small-N lab studies with
>    surprising outcomes, and anything that made a good headline** (§1.3 → `psych-epistemic-situation-and-replication-reform`).
> 3. **⚠️ Humans are far more shaped by situation and structure than intuition allows —
>    and far less infinitely malleable than a century of blank-slate theorizing
>    assumed.** **Both errors are common, and they're symmetrical.**

---

## §25. What's Live — verified August 2026

### 25.1 ⚠️ Where replication reform actually stands
**⚠️ The reforms work, and the field has not finished adopting them.**
- **⚠️ The strongest positive evidence remains the preregistration result** (§1.2 → `psych-epistemic-situation-and-replication-reform`):
  **four labs, 16 studies, 64 replications, 55 successes — an 86% replication rate**,
  against **the 30–70% typical of large-scale replication projects.**
- **⚠️ Recent estimates of the broader literature sit around half.** **A 2026 analysis
  reported roughly 50% of originally significant claims replicating; another put it at
  55.1%.** **Compare the Open Science Collaboration's 36% in 2015.**
- **⚠️ How to interpret ~50% is itself genuinely contested, and the debate is worth
  knowing.** **An optimistic reading: most original studies detected effects in the right
  direction and the problem is inflated effect sizes.** **A pessimistic reading: the
  central problems remain — implausibly high rates of significant results in journals,
  success rates inconsistent with typical statistical power, and repeated failures of
  headline findings.** ⚠️ **There's also a live argument that the "right" replication rate
  depends on your philosophy of science: a field consolidating phenomena should expect a
  higher rate than one aggressively testing risky theories.**
- **⚠️ A structural analysis argued that pre-reform parameter ranges IMPLY a replication
  rate of about 36% — matching what was observed** — **meaning a substantial part of the
  crisis was a predictable consequence of binary significance-based publishing rather
  than of bad researchers.**
- **⚠️ The reform is incomplete.** **Preregistration is still not the norm; reforms are
  not universally mandated; they do not retroactively fix the pre-2010s record; and
  incentives still favour positive novel results.** **One prominent critic's conclusion —
  that without enforceable field-wide standards, credibility remains a property of
  individual researchers rather than of the discipline — is harsh and hard to refute.**

**⚠️ Practical upshot, unchanged from §1.4 → `psych-epistemic-situation-and-replication-reform`**: **preregistered, large-N, independently
replicated, with reported effect sizes. Everything else is a hypothesis.**

### 25.2 ⚠️ Social media and adolescent mental health — a live dispute, presented as one
**⚠️ This is the most consequential contested empirical question in psychology right now,
it is genuinely unresolved, and I am going to give you both sides rather than pick.**
**It is also an unusually good case study in §2.4 → `psych-epistemic-situation-and-replication-reform` — what happens when a field cannot run
the experiment.**

**⚠️ The Haidt/Twenge position:**
- **Adolescent depression, anxiety and self-harm rose sharply in the US and several
  Western countries beginning in the early 2010s, coinciding with smartphone and social
  media adoption.**
- **⚠️ Cross-sectional studies number in the thousands and heavy users are consistently
  worse off**; **a 2022 meta-analysis reported odds of depression rising roughly 13% per
  additional daily hour of social media use.**
- **Longitudinal work (e.g. Riehm et al. 2019) has reported high use predicting roughly
  double the future risk of depression**, and **these findings were central to US Surgeon
  General warnings in 2023 and 2024.**
- **⚠️ The "blender study" critique**: **null results often pool all screen time and all
  demographics together; disaggregating by platform, sex and age reveals harms that
  aggregate analyses mask.**
- **Haidt and Rausch argue recent work — including experiments run by Meta itself —
  provides increasingly strong causal evidence, and that harms are occurring at
  population scale.**

**⚠️ The Odgers/Przybylski/Orben position:**
- **⚠️ Most of the evidence is correlational, and the effect sizes are very small.**
  **Odgers: hundreds of researchers have searched for the large effects claimed, and
  produced "a mix of no, small and mixed associations."**
- **⚠️ Reverse causation is at least as plausible as causation.** **Adolescents with
  existing mental health problems use these platforms differently and more** — **a
  Registered Report on nationally representative UK data (N=3,340) with clinical-rater
  diagnoses found exactly that pattern: adolescents with mental health conditions
  reported more social media time and less satisfaction with online friendships.**
- **⚠️ The international data don't fit well.** **An analysis across 72 countries found no
  consistent measurable association between wellbeing and social media rollout.**
- **⚠️ A selection-effect and moral-panic critique**: **the pattern resembles previous
  cycles about novels, radio, television and video games.**
- **⚠️ A pointed methodological objection**: **a review of the experimental studies cited
  as causal evidence found that almost all involved participants aged 18–35, mostly
  college students, and none involved children under 13 — while the policy prescriptions
  target children and young teens.**

**⚠️ And the dispute has become adversarial in a way that is itself informative.**
**Haidt and colleagues published a formal commentary in 2026 arguing three widely-cited
Vuorre and Przybylski papers should not be used to assess this question, on both
conceptual grounds and methodological ones — including Meta's involvement in the research
design of one study.** ⚠️ **Note that the conflict-of-interest concern cuts in a direction
worth registering: industry involvement in the null-result literature is a real issue, and
so is advocacy incentive in the harms literature.**

> **⚠️ GOTCHA — how to hold this honestly.** ⚠️ **"Correlation isn't causation" and "moral
> panic" are both true observations and neither is an argument that the harm is absent.**
> **Equally, a coincident time series and thousands of small correlations are not causal
> evidence.** **What would settle it — large, long, randomized, platform-specific trials
> in the actual target age group — is difficult, expensive and partly unethical, which is
> why it hasn't been done.**
> **⚠️ What both camps largely agree on**: **effects are heterogeneous rather than
> universal**; **girls and early adolescents appear more affected**; **what you do on a
> platform matters more than time spent**; **and displacement of sleep, exercise and
> in-person contact is a plausible mechanism independent of content.** ⚠️ **Anyone
> presenting this as settled in either direction is telling you about their priors.**

---

## §26. Misconceptions

| Misconception | Correction |
|---|---|
| Most published psychology findings are solid | ⚠️ **~36% replicated in 2015; ~50–55% in recent estimates** (§1 → `psych-epistemic-situation-and-replication-reform`, §25.1) |
| The replication crisis means psychology is fake | ⚠️ **Preregistered studies replicated at 86%** (§1.2 → `psych-epistemic-situation-and-replication-reform`) |
| **Learning styles** (visual/auditory/kinesthetic) | ⚠️ **No support. Repeatedly tested, repeatedly failed** (§1.3 → `psych-epistemic-situation-and-replication-reform`) |
| Ego depletion / willpower as a fuel tank | ⚠️ **Failed large multi-lab replication** (§7 → `psych-perception-memory-learning-cognition-and-emotion`) |
| Power posing changes hormones and outcomes | ⚠️ **Did not replicate** (§1.3 → `psych-epistemic-situation-and-replication-reform`) |
| Social priming effects are established | ⚠️ **Broad replication failure as a programme** (§12 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| The Stanford Prison Experiment shows situational power | ⚠️ **Methodologically compromised; not usable evidence** (§12 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| The IAT measures your personal bias | ⚠️ **Poor individual reliability and weak behaviour prediction** (§12 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| MBTI tells you your personality type | ⚠️ **Poor reliability, false dichotomies, no predictive validity** (§10 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| People are introverts or extraverts | ⚠️ **Continuous distribution, not types** (§10 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| Memory is a recording | ⚠️ **Reconstructive; false memories are readily created** (§5 → `psych-perception-memory-learning-cognition-and-emotion`) |
| Confident eyewitnesses are accurate | ⚠️ **Confidence tracks accuracy poorly, especially post-feedback** (§5 → `psych-perception-memory-learning-cognition-and-emotion`) |
| Re-reading and highlighting are good study methods | ⚠️ **Among the weakest. Use retrieval practice and spacing** (§5 → `psych-perception-memory-learning-cognition-and-emotion`) |
| We use 10% of our brains | ⚠️ **Straightforwardly false** |
| Left-brained / right-brained people | ⚠️ **Not a thing as popularly described** |
| It takes 21 days to form a habit | ⚠️ **No good basis; actual research found much longer and highly variable** (§6 → `psych-perception-memory-learning-cognition-and-emotion`) |
| Growth mindset produces large gains | ⚠️ **Effects far smaller than claimed** (§11 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| Different therapies differ greatly in effectiveness | ⚠️ **Broadly similar, except exposure for anxiety/OCD** (§13 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| Depression is a chemical imbalance | ⚠️ **Not supported as stated; medication can still work** (§13 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| Heritability means fixed or unchangeable | ⚠️ **Population variance statistic; Flynn effect disproves it** (§14 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| Heritability within groups explains gaps between groups | ⚠️ **Invalid inference. The most common error in the area** (§14 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| Multiple intelligences is established | ⚠️ **Popular; not supported psychometrically** (§14 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| Social construction means "not real" | ⚠️ **Constructed things have material consequences** (§16 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| Broken windows policing is evidence-based | ⚠️ **Much weaker support than its policy uptake** (§21 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| Severe punishment deters crime | ⚠️ **Certainty deters; severity much less** (§21 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| Modernization inevitably produces secularization | ⚠️ **Substantially revised; one regional trajectory** (§22 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| Hofstede's dimensions measure a person's culture | ⚠️ **Country-level aggregates; applying to individuals is an ecological fallacy** (§24 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| Psychology findings describe humans | ⚠️ **They mostly describe WEIRD samples** (§24 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| Social media's harm to teens is settled | ⚠️ **Genuinely contested. See both cases** (§25.2) |

---

## §27. Books

**⚠️ Epistemics first — read these before the content books:**
| Author | Work | Why |
|---|---|---|
| **Ritchie** | ***Science Fictions*** | ⚠️ **The best accessible account of §1–§3 → `psych-epistemic-situation-and-replication-reform`. Start here** |
| **Chivers** | *Everything Is Predictable* | Bayesian reasoning applied to exactly these problems |
| **Meehl** | *"Why Summaries of Research… Are Often Uninterpretable"* | ⚠️ **He diagnosed all of this in 1978. Chilling** |

**Psychology**
| **Kahneman** | *Thinking, Fast and Slow* | ⚠️ **Foundational AND partly superseded — read with §7 → `psych-perception-memory-learning-cognition-and-emotion` in hand** |
| **Barrett** | *How Emotions Are Made* | §8 → `psych-perception-memory-learning-cognition-and-emotion`'s constructionist case, argued strongly |
| **Loftus** | *Eyewitness Testimony* | §5 → `psych-perception-memory-learning-cognition-and-emotion` |
| **Nettle** | *Personality* | ⚠️ **Short, honest Big Five treatment** |
| **Brown, Roediger & McDaniel** | ***Make It Stick*** | ⚠️ **§5 → `psych-perception-memory-learning-cognition-and-emotion`'s learning findings, practically** |

**Sociology and anthropology**
| **Mills** | ***The Sociological Imagination*** | The founding statement of the discipline's move |
| **Goffman** | ***The Presentation of Self*** | ⚠️ **Still the most useful single sociology book** |
| **Bourdieu** | *Distinction* | §17–§18 → `psych-sociology-institutions-culture-and-the-weird-problem`. Difficult and worth it |
| **Granovetter** | *"The Strength of Weak Ties"* (1973) | ⚠️ **One paper, enormous payoff** (§19 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| **Geertz** | *The Interpretation of Cultures* | §23 → `psych-sociology-institutions-culture-and-the-weird-problem` |
| **Henrich** | ***The Secret of Our Success*** / *The WEIRDest People* | ⚠️ **§24 → `psych-sociology-institutions-culture-and-the-weird-problem`. The best cultural-evolution case** |
| **Scott** | *Seeing Like a State* | ⚠️ **Institutions, legibility, and why schemes fail** (§20 → `psych-sociology-institutions-culture-and-the-weird-problem`) |

---

## §28. Quick Reference

### 28.1 Picker
| Question | Where |
|---|---|
| Should I believe this psychology claim? | ⚠️ **§1.4 → `psych-epistemic-situation-and-replication-reform`'s checklist. Preregistered? Replicated? Effect size?** |
| Is this finding in the robust tier? | ⚠️ **§1.3 → `psych-epistemic-situation-and-replication-reform`** |
| How do I actually learn something? | ⚠️ **Retrieval practice + spacing** (§5 → `psych-perception-memory-learning-cognition-and-emotion`) |
| Which personality test? | ⚠️ **A validated Big Five inventory. Not MBTI** (§10 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| Does therapy work, and which? | ⚠️ **Yes; mostly alliance and therapist, except exposure** (§13 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| Is intelligence heritable / fixed? | ⚠️ **Heritable ≠ fixed. Flynn effect** (§14 → `psych-development-personality-social-clinical-and-neurodiversity`) |
| Why did this group outcome happen? | ⚠️ **Look at structure and networks before intentions** (§16 → `psych-sociology-institutions-culture-and-the-weird-problem`, §19 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| How do I find a job / novel information? | ⚠️ **Weak ties** (§19 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| Why do these organizations all do the same thing? | ⚠️ **Isomorphism** (§20 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| Does this generalize beyond the sample? | ⚠️ **Probably not — WEIRD** (§24 → `psych-sociology-institutions-culture-and-the-weird-problem`) |
| Is social media harming teenagers? | ⚠️ **Genuinely unsettled. §25.2** |

### 28.2 ⚠️ Red flags in any behavioural-science claim
- [ ] ⚠️ **Small sample with a surprising result**
- [ ] ⚠️ **Published pre-2015 and never independently replicated**
- [ ] ⚠️ **Effect size not reported, or reported only as p<.05**
- [ ] ⚠️ **A single study carrying a large policy or product recommendation**
- [ ] ⚠️ **Correlational data described with causal verbs**
- [ ] ⚠️ **A tidy mechanism story that outruns the data**
- [ ] ⚠️ **WEIRD sample, universal claim**
- [ ] ⚠️ **The researcher has a book, a consultancy, or a training programme built on it**
- [ ] ⚠️ **It's counterintuitive, memorable, and everyone has heard of it**

---

## §29. Method

**§1–§24 → `psych-epistemic-situation-and-replication-reform`, `psych-perception-memory-learning-cognition-and-emotion`, `psych-development-personality-social-clinical-and-neurodiversity`, `psych-sociology-institutions-culture-and-the-weird-problem` rest on stable material** — **the replication literature, the psychometric
findings, classical sociological theory, and the cross-cultural work** — sourced from
§27. ⚠️ **None of it needed verification.**

**Two searches were run in August 2026**, on **the current state of replication reform**
and **the social media/adolescent mental health dispute** — ⚠️ **the two places where a
static answer would misrepresent an active situation.**

**Confidence.** **High** in §1–§3 → `psych-epistemic-situation-and-replication-reform`, which I consider the most valuable part of this
document. ⚠️ **The structural decision to put epistemics FIRST rather than in an appendix
is deliberate: in this domain, knowing which findings survived is worth more than knowing
more findings**, and **most popular material about psychology draws on the pre-crisis
literature without flagging it.**

**High** in §1.2 → `psych-epistemic-situation-and-replication-reform`'s numbers — **the Open Science Collaboration's 36% and the preregistration
study's 86% (55 of 64 replications across four labs and 16 studies) are both well
documented** — ⚠️ **and I'd emphasise that the second is the more important and less
known.** **Recent broader estimates around 50–55% come from 2026 analyses, and I've
included the genuine disagreement about how to interpret that figure rather than
presenting one reading.**

**High** in the specific replication successes and failures listed in §1.3 → `psych-epistemic-situation-and-replication-reform`, §7 → `psych-perception-memory-learning-cognition-and-emotion`, §11 → `psych-development-personality-social-clinical-and-neurodiversity` and
§12 → `psych-development-personality-social-clinical-and-neurodiversity`. ⚠️ **These are individually well documented, and I've tried to be precise about
degree — "failed replication" for ego depletion and social priming, "much weaker than
claimed" for growth mindset and stereotype threat, "methodologically compromised" for
Stanford Prison.** **Those are different claims and collapsing them would be unfair in
both directions.**

⚠️ **§25.2 is presented as an unresolved dispute because it is one, and that is a
deliberate choice rather than fence-sitting.** **I've given each side its strongest form
with its actual evidence** — **the meta-analytic odds ratio and the disaggregation
critique on one side; the 72-country null, the Registered Report on reverse causation, and
the age-mismatch objection on the other.** ⚠️ **Both camps include serious researchers,
both have made valid criticisms of the other, and both operate under
conflict-of-interest pressures pointing opposite ways — industry involvement in some
null-result work, advocacy incentives in some harms work.** **I've flagged both.** **If
you want my honest read on where the burden sits: the population-level causal claim is not
established, the small-effects finding is well supported, and neither of those facts
means the harm is zero for the subgroups most affected.**

⚠️ **§14 → `psych-development-personality-social-clinical-and-neurodiversity` and §15 → `psych-development-personality-social-clinical-and-neurodiversity` were written with deliberate care** — **stating what is well established
and what is not, and specifically flagging the within-group/between-group heritability
inference as invalid, because that is where public discussion most often goes wrong in
both directions.**

⚠️ **Sourcing caution**: **the replication-reform material includes advocacy from both
reformers and defenders of the published record, and the social-media material includes
a great deal from partisans with books and public positions.** **Where I could anchor on
a Registered Report, a Nature or Science piece, or a large multi-lab project, I did.**