empirical-discourse / data.json
maximg1's picture
Upload folder using huggingface_hub
27f5f38 verified
Raw History Blame Contribute Delete
19.7 kB
{
"pending": {},
"corpus": {
"eras": {
"legacy": {
"n_sentences": 8443288,
"n_hearings": 11273,
"match_rate": 0.9874,
"congress_min": 105,
"congress_max": 114,
"parties": {
"Republican": 4839712,
"Democratic": 3485318,
"Independent": 11716
},
"female_share": 0.1407,
"by_congress": [
{
"congress": 105,
"year": 1997,
"n_sentences": 102810,
"n_hearings": 112
},
{
"congress": 106,
"year": 1999,
"n_sentences": 609141,
"n_hearings": 665
},
{
"congress": 107,
"year": 2001,
"n_sentences": 599023,
"n_hearings": 636
},
{
"congress": 108,
"year": 2003,
"n_sentences": 710447,
"n_hearings": 854
},
{
"congress": 109,
"year": 2005,
"n_sentences": 754593,
"n_hearings": 972
},
{
"congress": 110,
"year": 2007,
"n_sentences": 1279966,
"n_hearings": 1573
},
{
"congress": 111,
"year": 2009,
"n_sentences": 1037675,
"n_hearings": 1433
},
{
"congress": 112,
"year": 2011,
"n_sentences": 1254534,
"n_hearings": 1880
},
{
"congress": 113,
"year": 2013,
"n_sentences": 1088878,
"n_hearings": 1605
},
{
"congress": 114,
"year": 2015,
"n_sentences": 1006221,
"n_hearings": 1543
}
]
},
"modern": {
"n_sentences": 3616484,
"n_hearings": 4476,
"match_rate": 0.9893,
"congress_min": 115,
"congress_max": 118,
"parties": {
"Democratic": 1814154,
"Republican": 1763382,
"Independent": 80
},
"female_share": 0.2468,
"by_congress": [
{
"congress": 115,
"year": 2017,
"n_sentences": 838686,
"n_hearings": 1302
},
{
"congress": 116,
"year": 2019,
"n_sentences": 1103439,
"n_hearings": 1262
},
{
"congress": 117,
"year": 2021,
"n_sentences": 847792,
"n_hearings": 956
},
{
"congress": 118,
"year": 2023,
"n_sentences": 826567,
"n_hearings": 956
}
]
}
},
"by_congress": [
{
"congress": 105,
"year": 1997,
"n_sentences": 102810,
"n_hearings": 112,
"era": "legacy"
},
{
"congress": 106,
"year": 1999,
"n_sentences": 609141,
"n_hearings": 665,
"era": "legacy"
},
{
"congress": 107,
"year": 2001,
"n_sentences": 599023,
"n_hearings": 636,
"era": "legacy"
},
{
"congress": 108,
"year": 2003,
"n_sentences": 710447,
"n_hearings": 854,
"era": "legacy"
},
{
"congress": 109,
"year": 2005,
"n_sentences": 754593,
"n_hearings": 972,
"era": "legacy"
},
{
"congress": 110,
"year": 2007,
"n_sentences": 1279966,
"n_hearings": 1573,
"era": "legacy"
},
{
"congress": 111,
"year": 2009,
"n_sentences": 1037675,
"n_hearings": 1433,
"era": "legacy"
},
{
"congress": 112,
"year": 2011,
"n_sentences": 1254534,
"n_hearings": 1880,
"era": "legacy"
},
{
"congress": 113,
"year": 2013,
"n_sentences": 1088878,
"n_hearings": 1605,
"era": "legacy"
},
{
"congress": 114,
"year": 2015,
"n_sentences": 1006221,
"n_hearings": 1543,
"era": "legacy"
},
{
"congress": 115,
"year": 2017,
"n_sentences": 838686,
"n_hearings": 1302,
"era": "modern"
},
{
"congress": 116,
"year": 2019,
"n_sentences": 1103439,
"n_hearings": 1262,
"era": "modern"
},
{
"congress": 117,
"year": 2021,
"n_sentences": 847792,
"n_hearings": 956,
"era": "modern"
},
{
"congress": 118,
"year": 2023,
"n_sentences": 826567,
"n_hearings": 956,
"era": "modern"
}
],
"totals": {
"n_sentences": 12059772,
"n_hearings": 15749,
"match_rate": 0.988,
"congress_min": 105,
"congress_max": 118,
"year_min": 1997,
"year_max": 2023,
"year_end": 2024
},
"example": {
"speaker": "Ms. Tsongas",
"member": "Niki Tsongas",
"bioguide_id": "T000465",
"match_type": "hearing_member_exact",
"hearing": {
"title": "[H.A.S.C. No. 115-10] OVERVIEW OF MILITARY REVIEW BOARD AGENCIES",
"committee": "Committee on Armed Services",
"congress": 115,
"date": "2017-03-02"
},
"sentences": [
"Statutory deadlines requiring the BCMRs to complete 90 percent of their cases within 10 months create enormous pressure to move cases quickly.",
"Moreover, as the forces downsize, the expected caseloads for the boards may increase.",
"Yet over the past decade staffing for the boards has stayed the same or decreased.",
"What is the scope of the resource challenge that you face and how might Congress help?",
"Without increased staffing, how can you ensure that cases are adequately reviewed, records are collected, and that veterans' claims are given full and fair consideration?"
],
"covariates": {
"party": "Democratic",
"member_state": "Massachusetts",
"gender": "F",
"seniority": 6,
"nominate_dim1": -0.396,
"chairspeech": 0,
"minority": 1,
"vote_pct": 65.7309704435139
}
}
},
"labeling": {
"sets": [
{
"label": "Audit set",
"coders": "the original paper's coders",
"era": "1997-2016",
"n": 1500,
"pct_positive": 0.628
},
{
"label": "Modern gold",
"coders": "our two coders",
"era": "2017-2024",
"n": 1500,
"pct_positive": 0.554
}
],
"training": {
"label": "Training pool",
"coders": "both teams",
"era": "",
"n": 1504,
"pct_positive": 0.5944
},
"agreement": {
"kappa": 0.5913,
"n_overlap": 375
},
"benchmarks": [
{
"label": "Original labels",
"kappa": 0.44,
"pct_positive": 0.116
},
{
"label": "Audit set",
"kappa": 0.57
},
{
"label": "Modern gold",
"kappa": 0.5913
}
],
"adjudication": {
"n_disagreed": 76
},
"cases": [
{
"text": "You talked about training the trainers in your testimony, Dr. Redfield.",
"context_before": "They do wonderful work, but I am wondering if they are being fully brought into the messaging and the protection strategy.",
"context_after": "Maybe you could elaborate a little bit on that as well.",
"coder_a": 0,
"coder_b": 1,
"final": 0,
"category": null
},
{
"text": "And do you expect further support for the plan from any other organized unions?",
"context_before": "What has been the impact of the American Federation of State, County, and Municipal Employees' support for the plan of adjustment?",
"context_after": "nan",
"coder_a": 1,
"coder_b": 0,
"final": 0,
"category": null
},
{
"text": "What he did was he brought his own tool kit to school and told each kid to find something that is broken at your house, bring it in, and we will fix it as a class.",
"context_before": "Let me tell you what my dad did, and so many people have told me your dad taught me shop, or your dad, you know, was a great influence in my life.",
"context_after": "That had such a lasting impact on so many people, and has prepared the people that do my plumbing, and the people that fix my car, my dad taught them automotive shop.",
"coder_a": 0,
"coder_b": 1,
"final": 1,
"category": "descriptive"
},
{
"text": "We don't understand that trade-off between utility and privacy that is inherent in every one of us.",
"context_before": "Well, I mean, for me I think the steps that need to be taken are pretty straightforward.",
"context_after": "I've spent time in the Netherlands, and they have great open windows and--in big--in cities in the Netherlands, and that's when I realized that privacy is a different thing than anonymity, so that's one thing that we hav",
"coder_a": 0,
"coder_b": 1,
"final": 0,
"category": null
}
]
},
"models": {
"encoders": [
{
"encoder": "roberta",
"strategy": "inverse",
"auc_roc": 0.9128,
"macro_f1": 0.826,
"spread": 0.007
},
{
"encoder": "modernbert",
"strategy": "sqrt_inverse",
"auc_roc": 0.9039,
"macro_f1": 0.8219,
"spread": 0.0092
},
{
"encoder": "deberta",
"strategy": "inverse",
"auc_roc": 0.902,
"macro_f1": 0.8263,
"spread": 0.0053
}
],
"llama": [
{
"shots": 0,
"auc_roc": 0.8519,
"macro_f1": 0.5883
},
{
"shots": 3,
"auc_roc": 0.8611,
"macro_f1": 0.7243
},
{
"shots": 5,
"auc_roc": 0.8603,
"macro_f1": 0.7377
}
],
"ensembles": {
"single": {
"auc_roc": 0.8988,
"brier": 0.1263
},
"five_seeds": {
"auc_roc": 0.9143,
"brier": 0.1156
},
"full": {
"auc_roc": 0.9262,
"brier": 0.1067
}
},
"threshold": {
"value": 0.52,
"n_models": 30
}
},
"validation": {
"holdouts": {
"modern_test": {
"n": 748,
"metrics": {
"auc_roc": {
"legacy_only": 0.8691,
"mixed": 0.9117,
"delta": 0.0425,
"ci_low": 0.02,
"ci_high": 0.0681
},
"brier_score": {
"legacy_only": 0.162,
"mixed": 0.1366,
"delta": -0.0254,
"ci_low": -0.0458,
"ci_high": -0.007
},
"ece": {
"legacy_only": 0.1,
"mixed": 0.1093,
"delta": 0.0093,
"ci_low": -0.0498,
"ci_high": 0.0338
},
"cohens_kappa": {
"legacy_only": 0.5481,
"mixed": 0.5884,
"delta": 0.0403,
"ci_low": -0.0315,
"ci_high": 0.116
},
"macro_f1": {
"legacy_only": 0.773,
"mixed": 0.7941,
"delta": 0.021,
"ci_low": -0.0145,
"ci_high": 0.0597
}
}
},
"audit_gold": {
"n": 516,
"metrics": {
"auc_roc": {
"legacy_only": 0.921,
"mixed": 0.9164,
"delta": -0.0046,
"ci_low": -0.0178,
"ci_high": 0.0069
},
"brier_score": {
"legacy_only": 0.12,
"mixed": 0.135,
"delta": 0.015,
"ci_low": 0.0031,
"ci_high": 0.0274
},
"ece": {
"legacy_only": 0.0779,
"mixed": 0.1094,
"delta": 0.0315,
"ci_low": 0.0039,
"ci_high": 0.0514
},
"cohens_kappa": {
"legacy_only": 0.6517,
"mixed": 0.6368,
"delta": -0.0149,
"ci_low": -0.0599,
"ci_high": 0.0273
},
"macro_f1": {
"legacy_only": 0.8251,
"mixed": 0.8172,
"delta": -0.0079,
"ci_low": -0.0306,
"ci_high": 0.0134
}
}
}
},
"seam": {
"congress": 115,
"steps": [
{
"congress": 106,
"mixed": 0.00415,
"legacy_only": 0.00854
},
{
"congress": 107,
"mixed": -0.01335,
"legacy_only": -0.00932
},
{
"congress": 108,
"mixed": 0.00197,
"legacy_only": 0.00631
},
{
"congress": 109,
"mixed": -0.00205,
"legacy_only": -0.00565
},
{
"congress": 110,
"mixed": 0.00809,
"legacy_only": 0.00647
},
{
"congress": 111,
"mixed": -0.00425,
"legacy_only": 0.00147
},
{
"congress": 112,
"mixed": 0.00905,
"legacy_only": 0.00517
},
{
"congress": 113,
"mixed": -0.00478,
"legacy_only": -0.00705
},
{
"congress": 114,
"mixed": -0.00699,
"legacy_only": -0.00579
},
{
"congress": 115,
"mixed": -0.00597,
"legacy_only": -0.0004
},
{
"congress": 116,
"mixed": 0.00529,
"legacy_only": 0.00114
},
{
"congress": 117,
"mixed": -0.00421,
"legacy_only": 0.00251
},
{
"congress": 118,
"mixed": 0.0141,
"legacy_only": 0.01385
}
],
"placebo_p": 0.4615,
"diff_in_jumps": -0.00557,
"diff_ci_low": -0.00792,
"diff_ci_high": -0.00326,
"diff_placebo_p": 0.2308
}
},
"curated": {
"sentences": [
{
"text": "Sadly, by 2016, that number had risen to 44 percent.",
"p": 0.9966,
"human": 1,
"party": "D"
},
{
"text": "Mr. Chairman, I yield back.",
"p": 0.0044,
"human": 0,
"party": "D"
},
{
"text": "It is called an iron bamboo, I believe, is the species.",
"p": 0.5719,
"human": 0,
"party": "R"
},
{
"text": "This is what happens with limited competition.",
"p": 0.1076,
"human": 1,
"party": "R"
},
{
"text": "Twenty-two trillion dollars, I said it time and again on the campaign trail, and it resonated in my district.",
"p": 0.6126,
"human": 0,
"party": "R"
}
],
"stale": false,
"majority": "D",
"era": "116th Congress, 2019-20"
},
"finding": {
"trend": [
{
"congress": 105,
"year": 1997.0,
"majority": 0.25417,
"majority_ci_low": 0.22639,
"majority_ci_high": 0.27824,
"minority": 0.26486,
"minority_ci_low": 0.23987,
"minority_ci_high": 0.29383,
"gap": 0.01069,
"gap_ci_low": -0.02435,
"gap_ci_high": 0.05382,
"majority_party": "Republican"
},
{
"congress": 106,
"year": 1999.0,
"majority": 0.24741,
"majority_ci_low": 0.23308,
"majority_ci_high": 0.26441,
"minority": 0.29783,
"minority_ci_low": 0.28218,
"minority_ci_high": 0.31332,
"gap": 0.05042,
"gap_ci_low": 0.02799,
"gap_ci_high": 0.07187,
"majority_party": "Republican"
},
{
"congress": 107,
"year": 2001.0,
"majority": 0.22485,
"majority_ci_low": 0.2109,
"majority_ci_high": 0.24164,
"minority": 0.30029,
"minority_ci_low": 0.2889,
"minority_ci_high": 0.31268,
"gap": 0.07545,
"gap_ci_low": 0.05485,
"gap_ci_high": 0.09361,
"majority_party": "Republican"
},
{
"congress": 108,
"year": 2003.0,
"majority": 0.2252,
"majority_ci_low": 0.21125,
"majority_ci_high": 0.24134,
"minority": 0.29783,
"minority_ci_low": 0.28421,
"minority_ci_high": 0.31039,
"gap": 0.07263,
"gap_ci_low": 0.05212,
"gap_ci_high": 0.09077,
"majority_party": "Republican"
},
{
"congress": 109,
"year": 2005.0,
"majority": 0.22119,
"majority_ci_low": 0.20742,
"majority_ci_high": 0.2369,
"minority": 0.29537,
"minority_ci_low": 0.28176,
"minority_ci_high": 0.30801,
"gap": 0.07418,
"gap_ci_low": 0.05446,
"gap_ci_high": 0.09364,
"majority_party": "Republican"
},
{
"congress": 110,
"year": 2007.0,
"majority": 0.2433,
"majority_ci_low": 0.23106,
"majority_ci_high": 0.25578,
"minority": 0.28926,
"minority_ci_low": 0.27914,
"minority_ci_high": 0.30102,
"gap": 0.04596,
"gap_ci_low": 0.0295,
"gap_ci_high": 0.06348,
"majority_party": "Democratic"
},
{
"congress": 111,
"year": 2009.0,
"majority": 0.232,
"majority_ci_low": 0.21705,
"majority_ci_high": 0.24694,
"minority": 0.29851,
"minority_ci_low": 0.28832,
"minority_ci_high": 0.30856,
"gap": 0.0665,
"gap_ci_low": 0.04825,
"gap_ci_high": 0.08389,
"majority_party": "Democratic"
},
{
"congress": 112,
"year": 2011.0,
"majority": 0.24237,
"majority_ci_low": 0.22939,
"majority_ci_high": 0.25446,
"minority": 0.29884,
"minority_ci_low": 0.28555,
"minority_ci_high": 0.3108,
"gap": 0.05647,
"gap_ci_low": 0.03733,
"gap_ci_high": 0.07378,
"majority_party": "Republican"
},
{
"congress": 113,
"year": 2013.0,
"majority": 0.23482,
"majority_ci_low": 0.22481,
"majority_ci_high": 0.24617,
"minority": 0.30184,
"minority_ci_low": 0.28745,
"minority_ci_high": 0.31658,
"gap": 0.06702,
"gap_ci_low": 0.04857,
"gap_ci_high": 0.08446,
"majority_party": "Republican"
},
{
"congress": 114,
"year": 2015.0,
"majority": 0.23165,
"majority_ci_low": 0.22064,
"majority_ci_high": 0.24353,
"minority": 0.28987,
"minority_ci_low": 0.27626,
"minority_ci_high": 0.3035,
"gap": 0.05822,
"gap_ci_low": 0.04086,
"gap_ci_high": 0.07551,
"majority_party": "Republican"
},
{
"congress": 115,
"year": 2017.0,
"majority": 0.21879,
"majority_ci_low": 0.2077,
"majority_ci_high": 0.23078,
"minority": 0.28798,
"minority_ci_low": 0.27669,
"minority_ci_high": 0.29979,
"gap": 0.06919,
"gap_ci_low": 0.05165,
"gap_ci_high": 0.08547,
"majority_party": "Republican"
},
{
"congress": 116,
"year": 2019.0,
"majority": 0.22871,
"majority_ci_low": 0.21553,
"majority_ci_high": 0.24204,
"minority": 0.28409,
"minority_ci_low": 0.27277,
"minority_ci_high": 0.29521,
"gap": 0.05538,
"gap_ci_low": 0.03798,
"gap_ci_high": 0.0727,
"majority_party": "Democratic"
},
{
"congress": 117,
"year": 2021.0,
"majority": 0.20748,
"majority_ci_low": 0.19203,
"majority_ci_high": 0.2249,
"minority": 0.30328,
"minority_ci_low": 0.28968,
"minority_ci_high": 0.3201,
"gap": 0.09581,
"gap_ci_low": 0.07477,
"gap_ci_high": 0.11737,
"majority_party": "Democratic"
},
{
"congress": 118,
"year": 2023.0,
"majority": 0.24401,
"majority_ci_low": 0.2284,
"majority_ci_high": 0.25935,
"minority": 0.28842,
"minority_ci_low": 0.27631,
"minority_ci_high": 0.30025,
"gap": 0.04441,
"gap_ci_low": 0.0257,
"gap_ci_high": 0.06397,
"majority_party": "Republican"
}
],
"estimates": {
"mixed": {
"minority": {
"coef": 0.06062,
"se": 0.00354,
"ci_low": 0.05367,
"ci_high": 0.06756,
"p_value": 1.29e-65
},
"majority_mean": 0.2328,
"interaction": {
"coef": -0.00098,
"se": 0.00551,
"ci_low": -0.01178,
"ci_high": 0.00982,
"p_value": 0.859
}
},
"legacy_only": {
"minority": {
"coef": 0.04395,
"se": 0.00324,
"ci_low": 0.03759,
"ci_high": 0.0503,
"p_value": 8.06e-42
},
"majority_mean": 0.3126,
"interaction": {
"coef": -0.00028,
"se": 0.00497,
"ci_low": -0.01002,
"ci_high": 0.00946,
"p_value": 0.955
}
}
},
"flips": [
{
"year": 2007,
"before": 109,
"after": 110,
"parties": {
"Democratic": [
0.2981,
0.2577
],
"Republican": [
0.258,
0.2999
]
},
"effect": {
"coef": 0.04114,
"se": 0.00679,
"ci_low": 0.02783,
"ci_high": 0.05445,
"p_value": 1.38e-09
}
},
{
"year": 2019,
"before": 115,
"after": 116,
"parties": {
"Democratic": [
0.2879,
0.2503
],
"Republican": [
0.248,
0.2928
]
},
"effect": {
"coef": 0.04122,
"se": 0.00821,
"ci_low": 0.02513,
"ci_high": 0.05731,
"p_value": 5.15e-07
}
},
{
"year": 2023,
"before": 117,
"after": 118,
"parties": {
"Democratic": [
0.2585,
0.3
],
"Republican": [
0.2992,
0.2573
]
},
"effect": {
"coef": 0.04167,
"se": 0.008,
"ci_low": 0.026,
"ci_high": 0.05735,
"p_value": 1.87e-07
}
}
],
"measurement_error": {
"n_audit": 517.0,
"sensitivity_minority": 0.6941,
"sensitivity_majority": 0.7029,
"sensitivity_diff": -0.0088,
"sensitivity_diff_ci_low": -0.1171,
"sensitivity_diff_ci_high": 0.1067,
"corrected_score_gap": 0.1003,
"corrected_score_gap_ci_low": 0.0827,
"corrected_score_gap_ci_high": 0.1201
}
}
}