{"words as cues":{"train":{"words":[[1,3],[4,6],[3,1],[6,4],[1,3],[4,6],[1,3],[4,6],[3,1],[6,4],[2,3],[5,6],[3,1],[6,4],[2,3],[5,6],[3,2],[6,5],[3,2],[6,5],[3,1],[6,4],[2,3],[5,6],[3,2],[6,5],[3,2],[6,5]],"objects":[[1],[3],[1],[3],[1],[3],[1],[3],[1],[3],[2],[4],[1],[3],[2],[4],[2],[4],[2],[4],[1],[3],[2],[4],[2],[4],[2],[4]]},"test":[],"accuracy":[0.7115,"NA","NA",0.6731,"NA","NA"],"n_subj":[52],"label":["Kachergis2012-highlighting-words"],"condition":["words as cues (2 words, 1 object per trial)"],"description":["28 real training trials (2 replications x 7 early + 7 late), read from associative_word_learning/illcorr_hl2/orders/highlighting.txt (object indices relabeled to consecutive integers -- original objects 1/3/4/6 are objects 1/2/3/4 here, since 2 and 5 were unused foils). Early: {PE, I} jointly predict E (words 1&3 -> object 1; words 4&6 -> object 3); late: {PL, I} jointly predict L (words 2&3 -> object 2; words 5&6 -> object 4). Accuracy computed directly from the real N=67 test data (HLow_HLwo_67Ss.RData); PL and I items are NA -- see the limitation noted in data-raw/add_kachergis2012_highlighting.R. Kachergis, G. (2012). Learning Nouns with Domain-General Associative Learning Mechanisms. Proceedings of the 34th Annual Meeting of the Cognitive Science Society."],"response_matrix":{}},"objects as cues":{"train":{"words":[[1],[3],[1],[3],[1],[3],[1],[3],[1],[3],[2],[4],[1],[3],[2],[4],[2],[4],[2],[4],[1],[3],[2],[4],[2],[4],[2],[4]],"objects":[[1,5],[3,6],[5,1],[6,3],[1,5],[3,6],[1,5],[3,6],[5,1],[6,3],[2,5],[4,6],[5,1],[6,3],[2,5],[4,6],[5,2],[6,4],[5,2],[6,4],[5,1],[6,3],[2,5],[4,6],[5,2],[6,4],[5,2],[6,4]]},"test":[],"accuracy":[0.6951,0.6951,0.5,0.7317],"n_subj":[67],"label":["Kachergis2012-highlighting-objects"],"condition":["objects as cues (2 objects, 1 word per trial)"],"description":["The same 28-trial schedule as Kachergis2012-highlighting-words, with cue/outcome roles swapped: 2 objects + 1 word per trial (4 words, 6 objects; word and object indices both relabeled to consecutive integers and, for objects, so each word's own unique cue object shares its index -- original words 1/3/4/6 are words 1/2/3/4 here; original objects 1/2/4/5 (each word's own unique target) become 1/2/3/4, and the two shared/ambiguous objects (original 3, 6) become 5/6). Words 1 and 3 here (original 1, 4) are tested unambiguously (E, own target 1/3 here); words 2 and 4 (original 3, 6) are tested unambiguously (L, own target 2/4 here); the ambiguity is in the shared cue objects (5, 6 here), not in a tested word, so all 4 tested words get a real accuracy value, computed directly from the real N=67 test data (HLow_HLwo_67Ss.RData). Kachergis, G. (2012). Learning Nouns with Domain-General Associative Learning Mechanisms. Proceedings of the 34th Annual Meeting of the Cognitive Science Society."],"response_matrix":{}}}