Methods & evidence

A reader-facing audit of how strong the evidence base is: study designs, paper-level evidence strength, and claim-level grades.

1699

assessed papers

19.5%

causal-design share

16496

classified claims

Study design mix

Assessed papers by paper type.
Design	Papers
Descriptive	484
Theoretical	347
Quasi Experimental	238
Review Meta	233
Correlational	222
Rct	93
Commentary	55
Other	27

Paper-level evidence strength

Assessed papers by evidence-strength label.
Strength	Papers
Medium	818
N/a	419
Low	415
High	47

Claim-level evidence grades

Extracted claims by evidence grade.
Grade	Claims
Medium	8413
Low	2961
Speculative	2091
N/a	1545
High	1485
Medium-High	1

Highest-weighted evidence categories

Average evidence weight combines study type, evidence grade, and extraction confidence. Categories with fewer than five weighted claims are excluded.

Outcome categories with the highest average claim evidence weight.
Outcome category	Claims	Papers	Average evidence weight
Team Performance	176	83	0.263
Creative Output	70	30	0.261
Task Completion Time	272	194	0.258
Worker Satisfaction	220	115	0.249
Output Quality	896	347	0.247
Wages & Compensation	185	98	0.237
Decision Quality	740	319	0.228
Error Rate	210	128	0.228
Labor Share of Income	57	35	0.227
Innovation Output	401	210	0.211

Caveat

The causal-design share is a coarse screening label from automated assessment. It is useful for triage, but individual papers still need human review before making causal claims.