Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› devpsych-theory-and-hypotheses
devpsych-theory-and-hypotheses
GitHub专为发展心理学APA手稿设计，用于构建理论框架与年龄分级假设。明确变更机制、预测变化形式（如年龄主效应或交互作用），并严格区分确认性与探索性分析以符合JARS规范。
Trigger Scenarios
撰写理论框架和假设部分
准备发展研究的预注册或预分析计划
调和预测的变化轨迹与实际结果
回应审稿人关于缺乏理论或非发展性的批评
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill devpsych-theory-and-hypotheses -g -y
SKILL.md
Frontmatter
{
"name": "devpsych-theory-and-hypotheses",
"description": "Use when stating the developmental theory and age-graded hypotheses for a Developmental Psychology (APA) manuscript. The journal rewards hypotheses that are explicitly about change (age effects, within-person trajectories, mechanisms) and a clean confirmatory\/exploratory split under JARS. Structures the argument; it does not run analyses."
}
Developmental Theory & Hypotheses (devpsych-theory-and-hypotheses)
Developmental Psychology rewards a theory of change and age-graded, directional hypotheses. Because the journal asks for a credible developmental claim and applies JARS reporting, the cardinal rule is: state the developmental mechanism and the predicted form of change before the data, and label exploratory analyses as exploratory.
When to trigger
- Writing the theoretical framing and hypotheses
- Preparing a preregistration / pre-analysis plan for a developmental study
- Reconciling what you predicted about change with what the trajectories show
- A reviewer flagged the work as "atheoretical," "not developmental," "post hoc," or "HARKed"
Build the developmental argument
- State the mechanism of change. What developmental process (maturation, learning, transactional parent–child dynamics, scaffolding, cumulative risk) predicts the effect, and why it changes with age.
- Specify the form of the developmental hypothesis. Be explicit about which of these you predict: an age main effect, an age × condition interaction, a growth-curve shape (slope sign, acceleration), a within-person change, or a moderated trajectory.
- Derive directional predictions. Give a sign and, where possible, a magnitude/effect-size expectation — not "age will be related to X."
- Mark hypothesis status. Separate confirmatory (preregistered, predicted) from exploratory (generated after seeing data) analyses, clearly, per JARS.
- State what would disconfirm the developmental account — e.g., a flat trajectory, a reversed age slope, or invariance failure that means the construct is not the same across ages.
Age-graded hypothesis forms (name yours)
| Hypothesis form | What it claims | Typical test |
|---|---|---|
| Age main effect | level differs by age | age as predictor (cross-sectional or wave) |
| Age × condition | manipulation effect differs by age | interaction term |
| Growth shape | within-person slope/curvature | latent growth / multilevel |
| Sensitive period | effect concentrated in an age window | nonlinear / spline age terms |
| Transactional | child→parent and parent→child over time | cross-lagged / RI-CLPM |
Worked micro-example — theory to hypothesis status (illustrative)
A longitudinal effortful-control package, written so the developmental claim and status are legible.
Mechanism: Sensitive, contingent parenting scaffolds the maturation of
effortful control; the scaffolding payoff grows across early childhood.
H1 (confirmatory, preregistered): Effortful control shows positive
within-person growth from age 4 to 8 (latent slope > 0).
H2 (confirmatory): Maternal scaffolding at wave 1 predicts a steeper slope
(scaffolding × time interaction, directional).
H3 (exploratory): Bidirectional child→parent effects across waves
(RI-CLPM) — labeled exploratory; preregister for a future sample.
Disconfirming: a flat/negative latent slope, or measurement non-invariance
across ages (the construct is not the same thing being "grown").
Avoiding HARKing in developmental work
- Do not present an unexpected age interaction or a trajectory shape you discovered post hoc as if it had been predicted. If you found it after seeing the data, say so and frame it for confirmation.
- Constraints on generality matter here: name the ages, populations, and cohort/era to which the developmental claim is meant to apply, rather than implying it holds at every age and place.
Theory-stage reviewer pushback and the venue fix
| Reviewer pushback | Developmental Psychology fix |
|---|---|
| "Atheoretical / mechanism unclear" | state the developmental mechanism and why it changes with age in one or two sentences |
| "This is not developmental" | recast the hypothesis as change (slope/interaction/within-person), not a single-age correlation |
| "Looks HARKed" | show the preregistration; relabel post hoc trajectory shapes exploratory |
| "Claim outruns the design" | scale H1 to what the ages/waves can support; add a constraints-on-generality clause |
Anti-patterns
- Hypotheses about a relationship "in children/older adults" with no change component
- Non-directional "age will be associated with X" predictions
- Blurring confirmatory and exploratory analyses (JARS expects the split)
- Ignoring that an effect could be cohort, not age, and theorizing as if it were maturation
- No statement of what trajectory or interaction would disconfirm the account
Output format
【Mechanism of change】the developmental process, briefly
【Developmental hypothesis form】age effect / age×condition / growth shape / within-person / moderated
【Hypotheses】directional, testable (H1, H2, …)
【Status】confirmatory (preregistered) vs exploratory
【Disconfirming evidence】trajectory/interaction/invariance result that counts against it
【Next】devpsych-literature-positioning
Supplementary resources
../../resources/external_tools.md— preregistration templates and developmental-theory references../../resources/official-source-map.md— JARS reporting standards and scope
Version History
- 1839142 Current 2026-07-05 12:51
