Red for Danger, Warmth for Everything: How Children Assign Colours to Communicative Light Signals

This paper is a secondary, quantitative re-analysis of an open dataset of children's responses to the Firefly Design Task, in which children aged 3–15 assign colours to three imagined light signals (Glow, Hello, Danger), choose an animal to address, and write it a message. Rather than reporting colour preferences in the abstract, the study asks whether young children already reserve particular hues for particular communicative functions. Its central methodological choice is to treat each child's three colour responses as a within-subject (paired) design, using Cochran's Q and pairwise McNemar tests instead of pooling responses — avoiding the pseudoreplication common in colour-frequency studies. This separates two effects that raw frequencies conflate: a broad, non-specific preference for warm colours across all three signals, and a hue-specific reservation of red for the Danger signal. Secondary exploratory analyses examine engagement, message length across age, and gender differences in message content. The manuscript accompanies the underlying dataset deposit and is intended as a worked example of how a compact, well-documented open dataset can support rigorous developmental inference. It reports full test statistics, effect sizes, and confidence intervals, states its limitations plainly, and frames its secondary findings as hypotheses for confirmatory study. It will be of interest to researchers in colour cognition, early childhood development, and open/practitioner-led educational research, and to practitioners designing play-based science activities.