Table V. Correlation Between Root Mean Square Difference (RMSD) in Joint Reaction Forces From Scaled Models and the Discrepancies from the Underlying Datasets to the Modelled Subject.

	RMSDa
	Ankle		Knee		Hip		Sum
	R	p value	R	p value	R	p value	R	p value
∆genderb	0.354	0.001	0.373	<0.001	0.355	0.023	0.341	0.002
∆heightc	0.612	<0.001	0.585	<0.001	0.475	<0.001	0.528	<0.001
∆massc	0.662	<0.001	0.697	<0.001	0.463	<0.001	0.550	<0.001
∆limb lengthc	0.426	<0.001	0.702	<0.001	0.536	<0.001	0.435	<0.001
∆femur lengthc	0.426	<0.001	0.585	<0.001	0.520	<0.001	0.435	<0.001
∆tibia lengthc	0.490	<0.001	0.658	<0.001	0.536	<0.001	0.485	<0.001
∆pelvis widthc	0.103	0.335	0.101	0.354	0.256	0.001	0.142	0.212
∆BMIc	0.459	<0.001	0.360	0.001	0.401	<0.001	0.404	<0.001
∆limb length to pelvis width ratioc	0.324	0.002	0.441	<0.001	0.516	<0.001	0.380	0.001

The values in italics are statistically significant.

^aRMSDs in ankle, knee and hip joints, and the sum of these, are expressed as the percentage of mean force from the subject-specific model.

^bDifference in gender is defined as 1 or 0.

^cDifference in anthropometry measurements is expressed as a percentage difference from the underlying dataset to the subject; pelvis width is defined as the distance between right and left anterior superior iliac spine; the limb length to pelvis width ratio (ratio) is defined as limb length divided by pelvis width.