devpsych-tables-figures

Tables & Figures (devpsych-tables-figures)

In Developmental Psychology, the best exhibits show change. APA 7th-edition formatting is required,

but the venue-specific move is to make the developmental claim visible: age or time on the x-axis,

individual trajectories plus the model-implied mean curve, and uncertainty that a reader can see.

A bar chart of age-group means that hides within-person change is a missed opportunity here.

When to trigger

• Designing the main trajectory/results figure or a key table
• Deciding what goes in the article vs. online supplemental material
• A reviewer found an exhibit unclear, non-APA, or "doesn't show the developmental pattern"
• Visualizing growth curves, individual variation, and interactions with age

Principles

1. Make change visible. Prefer figures with age/time on the x-axis and the outcome on the y-axis:

spaghetti plots (individual trajectories) behind the model-implied mean curve, or growth curves by

moderator group. Show the developmental pattern, not just endpoint means.

1. Show uncertainty and effect size. Confidence bands/intervals on curves and estimates; report the

slope/interaction magnitude so the figure carries the claim, not just a p-value.

1. APA 7th + self-contained. Titles, notes, axis labels (with age units), Ns per age/wave, and what

the bands represent make each exhibit intelligible alone.

1. A measurement-invariance / model table earns its place. A compact table reporting invariance steps

(configural/metric/scalar fit, ΔCFI) reassures reviewers the change claim is sound.

1. Reproducible + accessible. Generated by the deposited analysis script so values match; colorblind-

safe and grayscale-legible.

Worked micro-example — the main trajectory figure (illustrative)

For the three-wave effortful-control package, the primary figure must show the growth, not bars.

Figure 1. Effortful-control trajectories, ages 4-8, by maternal scaffolding.
  Geometry:  thin per-child spaghetti lines + bold model-implied mean curves
             for high vs. low scaffolding; 95% CI bands on the mean curves.
  Axes:      x = child age in years (4-8); y = effortful-control score.
  Annotation: latent slope b = 0.42/yr [0.31, 0.53]; scaffolding × time
             b = 0.18 [0.07, 0.29].
  Note:      Ns per wave, attrition count, what the bands are (95% CIs),
             and that growth is interpreted under scalar invariance.
  Source:    rendered by the deposited R (lavaan/ggplot) script so values match.

Exhibit triage — article vs. online supplemental material

| Exhibit | Home | Reason |

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

| Main trajectory / growth figure with CIs | main text | this is the contribution |

| Measurement-invariance summary table | main text (compact) | underwrites the change claim |

| Full SEM path table / fit indices | supplement or appendix | secondary to the headline estimate |

| Attrition / completers-vs-dropouts table | supplement | credibility detail, not the headline |

| Robustness / alternative-model grid | supplement | summarized in one main-text sentence |

Exhibit-stage reviewer pushback and the venue fix

• "Bar chart of age groups hides the change" → switch to a trajectory/growth-curve figure with age on x.
• "I can't tell the developmental pattern" → overlay individual lines and the model-implied mean; label

the slope.

• "No uncertainty shown" → add CI bands and report the effect size in the note.
• "Figure values don't match the SEM table" → regenerate both from the single deposited script.

Exhibit calibration anchors

• The figure is where a hurried editor sees the developmental claim; design the main exhibit to stand

alone and to show change over age, not a pair of group means.

• Individual variation is part of the developmental story — show the spread of trajectories, not only the

average, especially when heterogeneity in change is theoretically interesting.

• A small invariance table is cheap insurance: it tells reviewers the construct is the same thing across

the ages you compare, pre-empting the most common developmental objection.

• Accessibility is part of credibility: colorblind-safe palettes and grayscale-legible encodings.

Anti-patterns

• Bar plots of age-group means that hide within-person change and variability
• Exhibits that need the prose to be intelligible (not self-contained)
• Trajectory figures with no confidence bands or effect sizes
• Omitting Ns per age/wave or the attrition count from notes
• Exhibit values that don't match the shared analysis script

Output format

【Main exhibit】what it shows + why a figure/table
【Shows change over age + uncertainty?】(trajectory/growth + CIs) [Y/N]
【Invariance table present?】[Y/N/NA]
【APA 7th + self-contained?】[Y/N]
【Article vs supplement】split decided
【Reproducible + accessible?】matches script, grayscale/colorblind-safe? [Y/N]
【Next】devpsych-writing-style

Supplementary resources

• [../../resources/external_tools.md](../../resources/external_tools.md) — ggplot2, tidySEM, papaja for APA exhibits, trajectory-plot tooling
• [../../resources/official-source-map.md](../../resources/official-source-map.md) — APA 7th style and exhibit requirements