Skip to main content
. Author manuscript; available in PMC: 2012 Oct 1.
Published in final edited form as: Neurobiol Aging. 2009 Dec 14;32(10):1742–1755. doi: 10.1016/j.neurobiolaging.2009.11.010

Table 6.

Cross-validation results for individual MCI subjects belonging to either the Progress group or the Stable group. A. Prediction summary for the cross-validation analysis including all 43 MCI subjects. T+ indicates a positive ERP test result (ERP test predicts MCI progression to AD) and T− indicates a negative ERP test result (ERP test predicts MCI remaining stable). Cut point for classification is 0.50. B. Prediction summary omitting the subjects that were “too close to call” based on their posterior probability of group membership. Cut point for classification is 0.70. The specificity, accuracy, and statistical significance increased when only the probable subjects were considered.

A.
MCI Subgroups T+ T− Total
Progress 10 5 15 Sensitivity: 0.67
Stable 8 20 28 Specificity: 0.71
Total 18 25 43 Accuracy: 70% χ2 = 5.82 (p=0.02)*
B.
MCI Subgroups T+ T− Total
Progress 9 4 13 Sensitivity: 0.69
Stable 3 16 19 Specificity: 0.84
Total 12 20 32 Accuracy: 78% χ2 = 9.41 (p=0.004)*
*

Corrected via Bonferroni’s adjustment for multiple comparisons.