Fig. 2
From: Quantifying spontaneous infant movements using state-space models
Changes in state occupancy during infant movement. (a) Individual movement dynamics were modelled using best-fit ARHMM. Top row: predicted motor states for each video frame in three videos chosen at random. Each frame is represented by a vertical line, coloured according to its predicted motor state. Middle row: Frames are sorted according to state membership. The time spent in each motor state varies across individuals. Bottom row. State occupancy (% frames in each state) for each video are shown as a bar plot. (b) Group average state occupancy. Proportion of time spent in each motor state is shown for all individual videos (n = 486, markers represent individual videos). Boxplot indicate median and interquartile range. (c) Change in state-occupancy with age was modelled using linear mixed effects models. Videos from the same individual are joined by a line. Black line indicate main effect of age with 95% confidence intervals shaded. Significant effects (p < 0.05 after Bonferroni correction over 8 states) are highlighted in bold.
