Let say you have 10 participants, and 38 channels. You extract one feature (alpha power) from each participant. Then your feature vector will be 10 x 38 (Instances x features). However, if you extract two features from each participant. Then your feature vector will be 10 x 76. Hope this clear.
Then for the label, you have 10 participant, let say first 5 of them are eye closed and last 5 of them are eye opened. Then your label will be [1;1;1;1;1;0;0;0;0;0]. The '1' means eye closed and the '0' means eye opened. The '1' and '0' you can change them with any value, but just make sure they are different since they represent different conditions (eye closed or opened).