Chi-square test values were calculated using two degrees of freedom. Data were categorized as follows: perpendicular (70–90°); planar (0–20°); oblique (20°–70°). As opposed to equal 30° bins, these weighted bins were chosen based on the observation that in a wild-type ‘bimodal’ division distribution, 80–90% of all divisions fell into the perpendicular and parallel categories. Thus a 20°–70° oblique bin better represents a deviation from normal than a 30°–60° bin would. All data were categorized in the same way. †Indicates that littermates were pooled from multiple hairpins for the same gene (for example, Pard3 1340 littermates and Pard3 1579 littermates), which is the actual data shown in the main text figure. This was done to avoid reproducing multiple control graphs in the main text. No statistical differences were ever found between age-matched littermates. ‡Denotes a specific cohort of ‘littermates’ associated with a single hairpin, in cases where the littermates data shown is pooled as above. §Indicates a data set which is not shown graphically in the main text figure panel.