A unique pattern of cortical connectivity characterizes patients with attention deficit disorders: a large electroencephalographic coherence study

Table 3 Classification of left-out clinical groups

1. Classification of medicated ADD (MED) on basis of CON vs. unmedicated ADD (UMED)
Training Set: CON vs. UMED				Test Set: MED
Num CON	% CON	Num UMED	% UMED	Num MED	% MED	DFA significance
correct	correct	correct	correct	correct	correct	t	df	P
550/619	88.9%	150/165	90.9%	161/182	88.46%	25.55	799	0.0001
Top 5 FAC: 13, 12, 4, 1, 2
2. Classification of ADD plus (APL) on basis of CON vs. ADD pure (APU)
Training Set: CON vs. APU				Test Set: APL
Num CON	% CON	Num APU	% APU	Num APL	% APL	DFA significance
correct	correct	correct	correct	correct	correct	t	df	P
556/619	89.8%	141/158	89.2%	164/189	86.77%	24.04	806	0.0001
Top 5 FAC: 13, 12, 4, 2, 3
3. Classification of ADD extreme (AEX) on basis of CON vs. ADD not extreme (nAEX)
Training Set: CON vs. nAEX				Test Set: AEX
Num CON	% CON	Num nAEX	% nAEX	Num AEX	% AEX	DFA significance
correct	correct	correct	correct	correct	correct	t	df	P
554/619	89.5%	219/238	90.2%	98/109	89.90%	20.33	726	0.0001
Top 5 FAC: 13, 12, 4, 2, 1

CON neurotypical controls, DFA discriminant function analysis, df degrees of freedom, FAC factor(s), t Student’s t-test, P probability

ISSN: 1741-7015