Skip to main content
. Author manuscript; available in PMC: 2009 Jun 29.
Published in final edited form as: J Neurosci Methods. 2007 Nov 19;168(2):431–442. doi: 10.1016/j.jneumeth.2007.11.003

Table II. Receiver-Operator Characteristics (ROC) analysis: Accuracy of SNAP components.

Accuracy of SNAP components. ROC analysis of each of the SNAP tests shows that the overall sum (SNAP score) has higher accuracy compared to its individual components. AUC represents overall accuracy with a perfect diagnostic test having an AUC of 1.0, and a test performing no better than chance having an AUC of 0.5. The p-value represents statistical significance compared with the null hypothesis, AUC = 0.5. (sham, n = 37; injured, n = 88).

Accuracy of SNAP components

Component AUC 95% Confidence Interval p value
Interactions 0.610 [0.498, 0.722] 0.0526
Cage grasp 0.658 [0.557, 0.759 0.0055
Visual placing 0.812 [0.737, 0.887] < 0.0001
Pacing or circling 0.502 [0.391, 0.614] 0.9698
Gait and posture 0.790 [0.710, 0.870] < 0.0001
Head tilt 0.722 [0.631, 0.813] < 0.0001
Visual field 0.8146 [0.738, 0.894] < 0.0001
Baton 0.891 [0.823, 0.958] < 0.0001

SNAP Score 0.971 [0.946, 0.996] < 0.0001