Regional Chinese folk-dance traditions, marked by complex symbolic gestures and diverse pedagogical legacies, face increasing threats of cultural dilution. This study aimed to quantitatively model and evaluate gesture transmission, recall, and instructional efficacy across three core Guangdong dance styles using high-rigour experimental protocols. A total of 48 participants were recruited from three instructional sites representing Cantonese Flower Drum, Hakka Rice Harvest, and Pearl River Boat dances. Data included 36 gesture units per style, captured with a dual-camera 4K video and analysed with Kinovea v0.9.5. Interventions consisted of stratified group assignment, spiral-loop instructional models, and audio-narrative integration. Outcomes assessed included gesture accuracy, motif integrity, recall, expressivity, angular drift, and motif gain using repeated measures ANOVA, two-way ANOVA, and post hoc Bonferroni tests. Pearl River Boat had the highest rhythmic density (3.6 ± 0.4 beats/unit) and peak motion angle (113.6 ± 6.9°), while Hakka Rice Harvest had the lowest angular range (101.1 ± 6.8°). Gesture ambiguity was most significant in Pearl River Boat (mean = 4.2 ± 0.8; 44.4% above threshold). Post-training recall was highest in Hakka Rice Harvest for motor accuracy (4.46 ± 0.36), and lowest in Pearl River Boat across recall domains. Inter-rater reliability was high (κ = .86 ± .03). Elder-to-youth instructional loops showed superior gesture accuracy (up to 6.51 ± 0.36) and the highest motif gain (51.9%). Quantitative pedagogical mapping can optimise folk dance curriculum and safeguard regional embodied heritage.
Weiyu Liu (Thu,) studied this question.