psychbull-theory-integration

Theory Integration (psychbull-theory-integration)

A Psychological Bulletin paper is judged not by the size of its table but by **what the synthesis

means for psychology. The journal publishes evaluative and integrative** reviews — the value is in

interpreting an existing literature, reconciling its contradictions, and advancing or refining a

theoretical account. (Note: a purely new theory with no synthesis belongs to Psychological Review;

here the theory is grounded in the synthesized evidence.) This skill builds the contribution; the

numbers come from the analysis skills.

When to trigger

• The pooled effect, moderators, and bias analyses exist and you must say what they mean
• A literature has conflicting findings you can now reconcile
• Building or refining a theoretical model from the synthesized evidence
• A reviewer says the review is "a competent meta-analysis but not a contribution"

From results to contribution

1. Answer the field's open question. State what the synthesis resolves: the size of an effect, why

studies disagreed, which moderator explains the pattern, whether an effect survives bias correction.

1. Reconcile conflict with the moderators. Use the moderator/meta-regression results to explain

why prior studies clashed — different populations, measures, designs, or conditions.

1. Build or refine a model. Translate the empirical pattern into a theoretical statement: an

integrative framework, boundary conditions, a mediating/moderating mechanism, or a corrected estimate.

1. State scope conditions honestly. Where does the effect hold, and where not? Heterogeneity and

the prediction interval discipline the claim.

1. Set the agenda. What should primary researchers do next — the gaps, the under-powered cells, the

untested moderators the synthesis exposes.

Calibrate claims to evidence

• Match the strength of the theoretical claim to heterogeneity, k, and bias robustness.
• Distinguish confirmatory (pre-specified) from exploratory patterns when theorizing.
• Avoid over-reading a fragile or bias-sensitive effect into a strong theory.

Anti-patterns

• A "results dump" with no interpretive or theoretical payoff
• Theorizing past the evidence (strong claims on small k or bias-sensitive effects)
• Ignoring heterogeneity when stating a general law
• Presenting an exploratory moderator as if it were the confirmatory story
• Submitting a pure new theory that should have gone to Psychological Review

The integration bar that defines a Bulletin paper

This is the calibration anchor that separates the APA's flagship review journal from a methods-only

meta-analysis: a competent pooled estimate is necessary but not sufficient. The genuine theoretical

advance is the threshold reviewers police.

| Synthesis delivers… | Methods-only meta-analysis (reject risk) | Psychological Bulletin contribution |

|---------------------|-------------------------------------------|--------------------------------------|

| A pooled number | "The effect is g = 0.34" and stops | Explains what that resolves for the field |

| Heterogeneity | Reports I² = 68% as a nuisance | Uses moderators to say why studies disagreed |

| A model | None — tallies effects | Advances or refines a theoretical account |

| Scope | Implies the effect is universal | States boundary conditions from the PI |

Worked vignette — turning numbers into theory

Illustrative numbers only. The self-affirmation synthesis yields g = 0.34 [0.24, 0.44], I² = 68%,

with delivery format explaining an R²-analog of 0.22 and bias diagnostics pulling the effect toward

≈ 0.25. The integrative payoff under this skill's rules:

• Open question resolved: self-affirmation has a small-to-moderate but real effect, smaller than the

early enthusiastic literature implied once selective reporting is modeled.

• Conflict reconciled: the disagreement tracks delivery format — facilitated, in-person formats

outperform brief written prompts — not a true-versus-null split.

• Model refined: this sharpens the timing/fit account rather than proposing a brand-new theory

(which would belong to Psychological Review).

• Scope conditions: the wide prediction interval [−0.10, 0.78] disciplines the claim — the effect can

vanish in some settings, so no universal law is asserted.

• Agenda: pre-registered trials varying delivery format, and direct tests of the underutilized cells.

Referee pushback → venue-specific fix

• "Competent meta-analysis, but not a contribution." → Add the interpretive layer: state the field

question and how the moderators resolve the conflict, not just the pooled number.

• "You theorize past your evidence." → Down-calibrate to k, heterogeneity, and bias robustness.
• "This is really a new theory." → If the model is not grounded in the synthesized evidence, it belongs

to Psychological Review.

Output format

【Open question resolved】one sentence
【Conflict reconciled】via which moderator(s)
【Model / framework】what is advanced or refined
【Scope conditions】where it holds / breaks
【Research agenda】top gaps for primary work
【Claim calibration】matched to k / heterogeneity / bias? [Y/N]
【Next】psychbull-tables-figures

Supplementary resources

• [../../resources/external_tools.md](../../resources/external_tools.md) — synthesis/reporting standards that frame the contribution
• [../../resources/official-source-map.md](../../resources/official-source-map.md) — Psychological Bulletin's evaluative/integrative scope (vs. Psychological Review for theory)