Skip to main content

Table 2 Although there is a significant correlation between most of the shape attributes and the two measures of the regression line, the percent variability “explained” by these relationships is generally fairly low, with all four attributes together accounting for only 4.1% and 20.2% of the variability in intercepts and slopes respectively

From: Evaluating the contribution of shape attributes to recognition using the minimal transient discrete cue protocol

  Intercept Slope
Attribute R R2 R R2
complexity 0.041 0.2% -0.368 13.6
mean curvature -0.132 1.7% -0.334 11.1%
inflection count -0.123 1.5% -0.310 9.6%
symmetry -0.143 2.1% 0.007 0.0%
Multiple R2   4.1%   20.0%