Insights
A meta-analysis: how the high-scoring cohort differs from the low-scoring / flagged one — by geography, impact and topic.
High scores (C1–C2)
643 publications87/100
mean reproducibility score
11
median citations
Geography (author institutions)
US
300
CN
116
DE
116
GB
94
FR
43
CA
35
AU
34
NL
33
Institutions
Harvard University
29
Universität Hamburg
28
University Medical Center Hamburg-Eppendorf
26
Centre National de la Recherche Scientifique
20
Inserm
17
Johns Hopkins University
14
University of Edinburgh
14
University of Pennsylvania
13
Top topics
Funders
no data yet
Impact tiers
very high: 12high: 52medium: 115low: 159
Low scores
192 publications50/100
mean reproducibility score
9
median citations
Geography (author institutions)
US
80
CN
59
DE
19
GB
14
AU
13
IT
11
CA
10
CH
8
Institutions
Harvard University
7
Cornell University
5
Inserm
5
Centre National de la Recherche Scientifique
5
Chinese Academy of Sciences
5
Universidad Nacional Autónoma de México
4
National Institutes of Health
4
Howard Hughes Medical Institute
4
Top topics
Funders
no data yet
Impact tiers
very high: 1high: 12medium: 36low: 52
What correlates with reproducibility?
Mean reproducibility score (0–100) across the assessed corpus, sliced by each dimension. Descriptive — correlation, not causation — and never a verdict on any group.
By impact tier
reproducibility vs. citation-based impact tier
very high
78/100
18
high
76/100
83
medium
74/100
215
low
73/100
302
By citation count
reproducibility vs. how often the paper is cited
0
71/100
91
1–9
74/100
211
10–99
75/100
274
100–999
76/100
40
1000+
84/100
2
By journal
highest- and lowest-scoring journals
Front Plant Sci
90/100
5
Microbiol Resour Announc
85/100
6
PLoS Pathog
85/100
6
Genome Med
83/100
9
Microb Genom
83/100
11
Sci Data
82/100
18
lowest-scoring
J Transl Med
60/100
4
Front Oncol
60/100
5
Cells
57/100
4
Mol Syst Biol
54/100
4
By country of the work
author-institution country (ISO codes)
GH
86/100
5
ZA
82/100
10
FI
81/100
16
TH
80/100
8
AR
80/100
6
LU
80/100
6
IN
79/100
34
IL
79/100
13
lowest-scoring
NO
64/100
15
IR
57/100
4
PT
57/100
5
HU
52/100
10
By institution
Vanderbilt University Medical Center
94/100
4
Boston VA Research Institute
94/100
4
University of Tartu
93/100
4
The University of Adelaide
93/100
6
Indian Institute of Science Bangalore
92/100
8
University Medical Center Freiburg
91/100
4
Centre de Coopération Internationale en Recherche Agronomique pour le Développement
90/100
5
University of Iowa
89/100
4
Data reuse vs. reproducibility
mean downstream reuse of the data accessions each cohort deposited
Reproduced (≥75)
106.2 papers/accession
50
Partial (50–74)
58.8 papers/accession
107
Discrepant (<50)
65 papers/accession
75
Each slice needs at least 4 assessed papers to appear, so a single paper can’t define a journal, country or funder. Bars: green ≥75 · amber ≥50 · red <50.
Descriptive patterns across what has been assessed so far — correlation, not causation, and sensitive to coverage. Not a judgement of any group or author.