Skills › Scientific Agent Skills › hypothesis-generation › falsification_controls_template.json
falsification_controls_template.json
116 lines · skills/hypothesis-generation/assets/falsification_controls_template.json at 5f2bc03
1{
2 "schema_version": "2.0",
3 "checklist_id": "FC-SYN-001",
4 "record_id": "HG-SYN-001",
5 "hypotheses": [
6 {
7 "hypothesis_id": "H1",
8 "candidate_status": "candidate",
9 "assumptions": [
10 "Synthetic assignment, manipulation, and measurements operate as declared."
11 ],
12 "boundary_conditions": [
13 "Synthetic eligible units over one prespecified interval."
14 ],
15 "falsifier": {
16 "prediction_id": "P1",
17 "conditions": "The manipulation and measurement checks pass and uncertainty is small enough to distinguish the patterns.",
18 "observable": "Intermediate M and outcome Y measured at prespecified times.",
19 "incompatible_result": "Outcome Y changes without the prespecified change in intermediate M.",
20 "assumption_failure_checks": [
21 "Verify assignment, manipulation, calibration, missingness, and temporal ordering before interpreting incompatibility."
22 ]
23 },
24 "discriminating_tests": [
25 {
26 "test_id": "DT1",
27 "rival_hypothesis_id": "H2",
28 "focal_expected": "Intermediate M changes before outcome Y with no negative-control reference shift.",
29 "rival_expected": "Outcome Y and the negative-control reference shift together without the prespecified intermediate change.",
30 "indeterminate_result": "Both intermediate M and the negative-control reference change, so mixed mechanisms or bias remain possible."
31 }
32 ],
33 "nulls": [
34 {
35 "null_id": "N1",
36 "statement": "The population mean of outcome Y is equal under X and comparator at the prespecified horizon.",
37 "analysis_id": "A1",
38 "interpretation_limit": "Failure to reject is not equivalence, absence of effect, or proof of the rival."
39 }
40 ],
41 "controls": [
42 {
43 "control_id": "NC1",
44 "control_type": "negative_outcome",
45 "rationale": "The synthetic reference cannot be changed by process M but shares the acquisition workflow.",
46 "expected_result": "No assignment-linked reference difference after the prespecified acquisition adjustment.",
47 "failure_implication": "Investigate acquisition, selection, or analysis bias before a causal interpretation."
48 },
49 {
50 "control_id": "PC1",
51 "control_type": "positive_control",
52 "rationale": "A known synthetic calibration input checks whether the outcome procedure detects a declared change.",
53 "expected_result": "The calibration input is recovered within the locked tolerance.",
54 "failure_implication": "The target test is not interpretable until measurement performance is restored."
55 }
56 ],
57 "outcome_interpretation": {
58 "consistent_with_candidate": "The focal pattern is observed, the rival pattern is not, controls pass, and uncertainty is compatible with the prespecified target.",
59 "challenges_candidate": "The incompatible result is observed while manipulation, measurements, and declared assumptions remain credible.",
60 "supports_neither_or_mixed": "The result matches neither pattern, both patterns occur, or an assumption/control fails."
61 },
62 "human_review_status": "pending"
63 },
64 {
65 "hypothesis_id": "H2",
66 "candidate_status": "candidate",
67 "assumptions": [
68 "Any synthetic acquisition drift affects the outcome and reference through a shared pathway."
69 ],
70 "boundary_conditions": [
71 "Synthetic acquisition periods covered by the declared reference process."
72 ],
73 "falsifier": {
74 "prediction_id": "P2",
75 "conditions": "Independent calibration periods overlap both assigned conditions and reference measurement is valid.",
76 "observable": "Synthetic negative-control reference and outcome Y by acquisition period.",
77 "incompatible_result": "The reference remains stable while outcome Y changes reproducibly by assignment.",
78 "assumption_failure_checks": [
79 "Verify reference sensitivity, batch overlap, calibration, and missing acquisition records."
80 ]
81 },
82 "discriminating_tests": [
83 {
84 "test_id": "DT2",
85 "rival_hypothesis_id": "H1",
86 "focal_expected": "Reference and outcome shift together by acquisition period.",
87 "rival_expected": "Outcome changes by assignment without a reference shift and after intermediate M changes.",
88 "indeterminate_result": "Acquisition period is confounded with assignment, so the rivals cannot be separated."
89 }
90 ],
91 "nulls": [
92 {
93 "null_id": "N2",
94 "statement": "The negative-control reference mean is equal across acquisition periods.",
95 "analysis_id": "A2",
96 "interpretation_limit": "Failure to reject does not prove absence of drift when precision or reference sensitivity is insufficient."
97 }
98 ],
99 "controls": [
100 {
101 "control_id": "NC1",
102 "control_type": "negative_outcome",
103 "rationale": "The reference is outside process M while sharing the acquisition pathway.",
104 "expected_result": "A drift explanation predicts a shared acquisition-linked shift.",
105 "failure_implication": "A stable, sensitive reference challenges the drift candidate."
106 }
107 ],
108 "outcome_interpretation": {
109 "consistent_with_candidate": "Outcome and reference shifts align by acquisition period under valid measurement.",
110 "challenges_candidate": "The stable reference and independent calibration are incompatible with the declared shared-drift process.",
111 "supports_neither_or_mixed": "Assignment and acquisition cannot be separated or both mechanism and drift patterns occur."
112 },
113 "human_review_status": "pending"
114 }
115 ]
116}