Skip to main content
. Author manuscript; available in PMC: 2024 Feb 19.
Published in final edited form as: Neuroimage. 2022 Jul 14;260:119457. doi: 10.1016/j.neuroimage.2022.119457

Fig. 2.

Fig. 2.

Results of machine learning analysis. Each panel shows the likelihood of the classifier choosing the target (the word that was heard, e.g., badger), its cohort (an onset competitor, e.g., baggage) or an unrelated word (e.g., mushroom) as a function of time. Word onset is at 0 sec. Each point marks the onset of a bin (e.g., data at .25 sec represents a classification analysis using EEG data from .250 to .500 sec). Chance (thin gray line) is 0.125. A) Results for classification among the 8 words in subjects tested on a 64-channel low impedance EEG at Iowa (N = 16). B) Results for classification among the 8 nonwords in the Iowa sample. C) Results for classification among the words for subjects tested in a 128-channel high impedance EEG at Haskins (N = 15). D) Results for classification among nonwords for Haskins sample.