{"success":true,"database":"eegdash","data":{"_id":"6953f4249276ef1ee07a3321","dataset_id":"ds004147","associated_paper_doi":null,"authors":["Cameron D. Hassall","Laurence T. Hunt","Clay B. Holroyd"],"bids_version":"1.2.1","contact_info":["Cameron Hassall"],"contributing_labs":null,"data_processed":false,"dataset_doi":"doi:10.18112/openneuro.ds004147.v1.0.2","datatypes":["eeg"],"demographics":{"subjects_count":12,"ages":[24,30,22,24,23,52,56,32,42,77,68,46],"age_min":22,"age_max":77,"age_mean":41.333333333333336,"species":null,"sex_distribution":{"f":7,"m":5},"handedness_distribution":{"r":11,"l":1}},"experimental_modalities":null,"external_links":{"source_url":"https://openneuro.org/datasets/ds004147","osf_url":null,"github_url":null,"paper_url":null},"funding":[],"ingestion_fingerprint":"761ccec29b3b84dc0ed100790110a0c07cc58728fab9a2772cd4c9b06c59c3c3","license":"CC0","n_contributing_labs":null,"name":"Average Task Value","readme":"### Average Task Value\nTwelve participants completed three learning tasks. In each task the goal was to learn cue-response mappings for six cues. The cues were various coloured shapes. The possible responses were left ('d' key) or right ('k' key). There were two types of cues. Low-value cues had a feedback validity of 0.5 (i.e., a coin toss). High-value cues had a feedback validity of 0.8 (80% chance of a win if the correct action was chosen). The low-value task contained only low-value cues. The high-value task contained only high-value cues. The mid-value task contained three low-value cues and three high-value cues. Participants completed 144 trials of each task.\nPreprint: https://doi.org/10.1101/2021.09.16.460600\nAnalysis code: https://github.com/chassall/averagetaskvalue","recording_modality":["eeg"],"senior_author":"Clay B. Holroyd","sessions":[],"size_bytes":4291179548,"source":"openneuro","study_design":null,"study_domain":null,"tasks":["casinos"],"timestamps":{"digested_at":"2026-04-22T12:26:09.920200+00:00","dataset_created_at":"2022-06-07T14:19:20.839Z","dataset_modified_at":"2024-01-24T19:53:26.000Z"},"total_files":12,"storage":{"backend":"s3","base":"s3://openneuro.org/ds004147","raw_key":"dataset_description.json","dep_keys":["CHANGES","README","participants.json","participants.tsv"]},"tagger_meta":{"config_hash":"4a051be509a0e3d0","metadata_hash":"65144ddd933d2a5d","model":"openai/gpt-5.2","tagged_at":"2026-01-20T10:32:27.084551+00:00"},"tags":{"pathology":["Healthy"],"modality":["Visual"],"type":["Learning"],"confidence":{"pathology":0.6,"modality":0.8,"type":0.8},"reasoning":{"few_shot_analysis":"Most similar few-shot example by paradigm is \"EEG: Probabilistic Learning with Affective Feedback: Exp #2\" (Healthy, Visual, Learning): both involve reinforcement/probabilistic feedback to learn cue-response/action mappings using visual cues. That example guides mapping this dataset to Type=Learning (reinforcement learning paradigm) and Modality=Visual (screen-based cues).","metadata_analysis":"Key facts from the provided README: (1) Task/construct: \"Twelve participants completed three learning tasks\" and \"the goal was to learn cue-response mappings for six cues.\" (2) Stimulus modality: \"The cues were various coloured shapes.\" (3) Reinforcement/probabilistic feedback structure: \"Low-value cues had a feedback validity of 0.5\" and \"High-value cues had a feedback validity of 0.8 (80% chance of a win if the correct action was chosen).\" (4) Response is keypress but does not define modality: \"possible responses were left ('d' key) or right ('k' key).\" No clinical recruitment or diagnosis is mentioned.","paper_abstract_analysis":"No useful paper information (only a preprint link is provided in the README, but no abstract text is included here).","evidence_alignment_check":"Pathology: Metadata SAYS participants did learning tasks but gives no diagnosis/clinical recruitment (e.g., \"Twelve participants completed three learning tasks\"), so pathology must be inferred as non-clinical. Few-shot pattern SUGGESTS Healthy for typical lab learning tasks without patient labels. ALIGN (no conflict).\nModality: Metadata SAYS stimuli are visual (\"cues were various coloured shapes\"). Few-shot pattern for similar reinforcement-learning tasks also uses Visual. ALIGN.\nType: Metadata SAYS explicit learning aim (\"learning tasks\"; \"goal was to learn cue-response mappings\"). Few-shot convention labels similar probabilistic feedback mapping tasks as Learning (e.g., probabilistic learning dataset). ALIGN.","decision_summary":"Top-2 candidates per category:\n- Pathology: (1) Healthy — supported by absence of any clinical population description and generic \"participants\" wording; (2) Unknown — because README does not explicitly say \"healthy\" or \"controls\". Winner: Healthy. Alignment: aligned with few-shot convention; no metadata conflict.\n- Modality: (1) Visual — \"cues were various coloured shapes\"; (2) Other — if cues were not presented visually (unlikely). Winner: Visual. Alignment: aligned.\n- Type: (1) Learning — \"learning tasks\" and \"goal was to learn cue-response mappings\" with probabilistic feedback; (2) Decision-making — because participants choose left/right actions with reward outcomes, but primary described purpose is learning mappings. Winner: Learning. Alignment: aligned.\nConfidence basis: Modality and Type have multiple explicit supporting quotes; Pathology lacks explicit health-status statement and is inferred from absence of diagnosis."}},"nemar_citation_count":2,"computed_title":"Average Task Value","nchans_counts":[{"val":31,"count":12}],"sfreq_counts":[{"val":1000.0,"count":12}],"stats_computed_at":"2026-04-22T23:16:00.307124+00:00","total_duration_s":null,"author_year":"Hassall2022_Average","canonical_name":null}}