Verified Rigor RecordRigor Reviewed

Rigor-Reviewed Manuscript

miR-34a Blocks Osteoporosis and Bone Metastasis by Inhibiting Osteoclastogenesis and Tgif2 · 10.1038/nature13375

Name: Verified Rigor Record — miR-34a Blocks Osteoporosis and Bone Metastasis by Inhibiting Osteoclastogenesis and Tgif2 · 10.1038/nature13375
Item: miR-34a Blocks Osteoporosis and Bone Metastasis by Inhibiting Osteoclastogenesis and Tgif2 · 10.1038/nature13375
Rating: 75
Author: Αlpha¹

Paper source

Read full text

Αlpha¹ Editorial

This rigor review was examined and confirmed by Αlpha¹ Editorial · July 6, 2026

Αlpha¹ Rigor Review

Rigor Reviewed

WARN

Warnings to address · Interventional Study

Eight dimensions — click a segment for details

Critical

High-priority

Model-reported confidence: moderateReviewer agreement: lowEngine: 7.6.0Source: full text

Not verified in this report

No reported statistical tests were found to recompute.
No data or code availability links were detected to verify.

This Αlpha¹ Rigor Review uses AI Rigor Reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.

The paper presents a well-motivated study with a clear scientific premise and detailed methods, but has significant reporting gaps in study design (randomization, blinding, replicate distinction), key resources (RRIDs, cell line authentication, mycoplasma testing), statistical analysis (exact p-values, effect sizes, assumptions), and data/code availability. These gaps reduce reproducibility and transparency.

Three independent reviewers evaluated the full text. The citation check found no retracted or missing references. The statistics component could not verify any tests due to lack of reported test statistics with degrees of freedom. The preregistration check found no registration. The claim audit found no under-evidenced claims.

8 major claims checked against the paper's own evidence: all adequately supported.

1 integrity concern flagged (0 high).

Rigor

Eight Dimensions5 pass · 3 fail

8 major claims1 partial7 supported

Prior work is cited for miRNAs in physiology and disease, and the study builds on the underexplored role of miRNAs in skeletal biology. The rationale logically links the observed downregulation of miR-34a during osteoclast differentiation to the hypothesis that miR-34a is a suppressor of osteoclastogenesis and bone resorption. Limitations of prior work (e.g., lack of genetic evidence) are addressed through the use of multiple genetic models.

Sub-criteria

prior work citedADEQUATE

premise rationaleADEQUATE

limitations addressedADEQUATE

Per reviewer (3) — the votes behind the verdict

PASSReviewer 1· 90% conf

The manuscript establishes a clear scientific premise by citing prior research on miRNAs in bone and providing a logical rationale for studying miR-34a, though it does not deeply critique limitations of earlier studies.

Evidence

direct quote[Introduction]

“MicroRNAs (miRNAs) play important roles in physiology and disease , , and present tremendous therapeutic potential . Nonetheless, how miRNAs regulate skeletal biology is underexplored.”

direct quote[Results, paragraph 1]

“miR-34a was rapidly down-regulated by RANKL and further diminished by rosiglitazone”

absence[Introduction]

Limitations of prior work are not critically discussed.

PASSReviewer 2· 95% conf

The study's premise is well-established through a clear rationale from prior work on miRNAs in skeletal biology and the identification of miR-34a as a downregulated miRNA during osteoclastogenesis.

Evidence

direct quote[Abstract, Introduction (paragraph 1)]

“MicroRNAs (miRNAs) play important roles in physiology and disease and present tremendous therapeutic potential. Nonetheless, how miRNAs regulate skeletal biology is underexplored.”

direct quote[Results, paragraph 1]

“miR-34a was rapidly down-regulated by RANKL and further diminished by rosiglitazone”

direct quote[Discussion, paragraph 1]

“The roles of miRNAs in bone physiology have just begun to emerge.”

PASSReviewer 3· 95% conf

The paper clearly establishes the scientific premise by citing prior work on miRNAs in bone biology, the role of osteoclasts in osteoporosis and bone metastasis, and the therapeutic potential of miRNAs.

Evidence

direct quote[Abstract, Introduction (first paragraph)]

“MicroRNAs (miRNAs) play important roles in physiology and disease, and present tremendous therapeutic potential. Nonetheless, how miRNAs regulate skeletal biology is underexplored.”

direct quote[Abstract]

“Here we identify miR-34a as a novel and critical suppressor of osteoclastogenesis, bone resorption and the bone metastatic niche.”

paraphrase[Abstract and throughout Results]

“Paraphrase: The paper uses 'mouse genetic, pharmacological and disease models' to reveal miR-34a as a key osteoclast suppressor, moving beyond correlative expression studies.”

Randomization and blinding are reported for analyses. Power analysis is described with a specific example. Inclusion/exclusion criteria are stated (no animals excluded). Outlier handling is not reported. Controls are present (e.g., sham, littermate, miR-Ctrl). Independent replication is achieved through multiple independent founders and cre drivers. Replicate distinction (biological vs technical) is not explicitly defined. Randomization method is stated as 'Samples were randomly allocated' but the specific method is not detailed, so it is rated as reported_but_inadequate.

Sub-criteria

controlsADEQUATE

power analysisADEQUATE

blinding levelsADEQUATE

outlier handlingNOT REPORTED

randomization unitADEQUATE

inclusion exclusionADEQUATE

randomization methodINADEQUATE

replicate distinctionNOT REPORTED

independent replicationADEQUATE

Per reviewer (3) — the votes behind the verdict

WARNReviewer 1· 85% conf

The study design includes appropriate controls and replication across multiple models, but randomization method, unit, and blinding are incompletely reported, and replicate types are not distinguished.

Evidence

direct quote[Methods]

“Samples were randomly allocated to each group.”

direct quote[Methods]

“Sample size estimate was based on power analyses performed using SAS 9.3 TS X64_7PRO platform. With the observed group differences which is of great biological value, and the relatively small variation of the in vivo measurements, a sample size of four per group (n=4) will provide higher than 90% power at type I error rate of 0.05 (two-sided test), and a sample size of three per group (n=3) will provide higher than 80% power at type I error rate of 0.05 (two-sided test).”

direct quote[Methods]

“Analyses were conducted in a blind fashion to the operator.”

PASSReviewer 2· 85% conf

The study design is rigorous, with multiple mouse models, randomization, blinding, and sample-size estimates, though a few criteria are not fully reported or are not applicable.

Evidence

direct quote[Methods, paragraph 6 (Mice)]

“Sample size estimate was based on power analyses performed using SAS 9.3 TS X64_7PRO platform.”

direct quote[Methods, paragraph 7 (Mice)]

“Analyses were conducted in a blind fashion to the operator.”

direct quote[Statistical Analyses, Methods final paragraph]

“No animal or sample was excluded from the analysis.”

PASSReviewer 3· 80% conf

The study uses a comprehensive design with multiple transgenic models, power analysis, blinding, and randomization, though blinding details for some assays and independent replication for some experiments could be clearer.

Evidence

direct quote[Methods (Mice section, final paragraph)]

“Samples were randomly allocated to each group. Analyses were conducted in a blind fashion to the operator.”

paraphrase[Methods (Mice section) and Results (Extended Data 2, 3)]

“Paraphrase: The paper states 'Representative results from at least two independent founders are reported here' and uses four different osteoclast-targeting cre drivers (Tie2, PT, Lys, Ctsk) to confirm the miR-34a transgenic phenotype.”

direct quote[Methods (Statistical Analyses)]

“No animal or sample was excluded from the analysis.”

Sex is reported in figure legends (e.g., '2-month-old, male'). Age is consistently reported. Strain/source is reported (C57BL/6J, mixed backgrounds). Housing conditions are not explicitly described. Health status is not explicitly stated but implied by specific-pathogen-free conditions. Demographics (human) are not applicable as the study uses mouse models. Sex justification is not applicable as both male and female mice are used in some experiments.

Sub-criteria

demographicsN/A

sex reportedADEQUATE

sex justifiedN/A

age weight healthADEQUATE

housing conditionsNOT REPORTED

species strain sourceADEQUATE

Per reviewer (3) — the votes behind the verdict

WARNReviewer 1· 80% conf

Sex and strain are reported, but housing conditions are not described, age is reported but health status is not, and single-sex experiments lack justification.

Evidence

direct quote[Methods/Results]

“2-month-old, male, n=4”

absence[Methods]

Housing conditions not reported.

direct quote[Methods, Mice]

“mice on pure C57BL/6J background, and compared to littermate controls that carry only the transgene allele or only the Cre allele”

PASSReviewer 2· 95% conf

Sex, age, and mouse strain are well-reported for the key experiments, though some details like health status and housing conditions are partially reported.

Evidence

direct quote[Figure 2 legend, panels d-f]

“2-month-old, male, n=4”

direct quote[Methods, Mice]

“transgenic founders on pure C57BL/6J background”

absence[Methods, Mice]

Housing conditions are not described.

PASSReviewer 3· 95% conf

Sex, age, strain, source, and health status are reported for all mouse experiments. Housing conditions are not described.

Evidence

direct quote[Figure legends (e.g., Figure 2c, 2e, 3, etc.)]

“c, Serum CTX-1 (2-month-old, male, n=5).”

direct quote[Methods (Mice section)]

“Conditional miR-34a transgenic mice were generated using the CAG-Z-EGFP vector... Transgenic founders on pure C57BL/6J background were established by pronuclear injection at the UT Southwestern transgenic core.”

absence[Methods (Mice section)]

Housing conditions are not reported.

The study uses mice and reports IACUC approval from the named institution. Regulatory compliance is stated through adherence to the IACUC guidelines. The IACUC statement is adequate as it names the institution and committee.

Sub-criteria

iacuc statementADEQUATE

informed consentN/A

irb ethics statementN/A

regulatory complianceADEQUATE

Per reviewer (3) — the votes behind the verdict

WARNReviewer 1· 85% conf

IACUC approval from UT Southwestern is stated, but compliance with a specific animal care regulation (e.g., Guide for the Care and Use of Laboratory Animals) is not explicitly mentioned.

Evidence

direct quote[Methods, Mice]

“All protocols for mouse experiments were approved by the Institutional Animal Care and Use Committee of University of Texas Southwestern Medical Center.”

direct quote[Methods]

“No statement of adherence to a specific regulation such as the Guide for the Care and Use of Laboratory Animals.”

PASSReviewer 2· 95% conf

Animal ethics approval is reported with a named institutional committee and regulatory compliance.

Evidence

direct quote[Methods, Mice, final sentence]

“All protocols for mouse experiments were approved by the Institutional Animal Care and Use Committee of University of Texas Southwestern Medical Center.”

direct quote[Methods, Mice]

“All protocols for mouse experiments were approved by the Institutional Animal Care and Use Committee of University of Texas Southwestern Medical Center.”

PASSReviewer 3· 95% conf

Animal ethics approval is reported with a named IACUC body and institutional location, and regulatory compliance is stated.

Evidence

direct quote[Methods (Mice section, final paragraph)]

“All protocols for mouse experiments were approved by the Institutional Animal Care and Use Committee of University of Texas Southwestern Medical Center.”

Antibodies are listed with vendor and catalog but no RRIDs or validation. RAW264.7 is identified by ATCC number but not authenticated. Mycoplasma testing not mentioned. Reagents often lack catalog numbers. Software version for μCT not given.

Sub-criteria

mycoplasma testingNOT REPORTED

reagents identifiedINADEQUATE

organisms identifiedADEQUATE

antibodies identifiedINADEQUATE

cell line authenticationINADEQUATE

software tools identifiedINADEQUATE

Per reviewer (3) — the votes behind the verdict

FAILReviewer 1· 90% conf

Only mouse strains are adequately identified; antibodies, cell lines, reagents, and software lack complete identification (RRIDs, catalog numbers, versions), and mycoplasma testing is not reported.

Evidence

direct quote[Methods, Reagents]

“Anti-Tgif2 antibody (catalog # 09-718) was from Millipore; anti-NFATc1 (7A6), anti c-Jun (H-79) anti IκBα (C-21) antibodies were from Santa Cruz Biotechnologies”

direct quote[Methods, Reagents]

“RAW264.7 mouse macrophage cell line was from ATCC (TIB-71).”

absence[Methods]

Mycoplasma testing not reported.

PASSReviewer 2· 90% conf

Key reagents and software are well-identified, including antibodies, cell lines, and transgenic mice, with vendors and catalog numbers provided.

Evidence

direct quote[Methods, Reagents]

“Anti-Tgif2 antibody (catalog # 09-718) was from Millipore; anti-NFATc1 (7A6), anti c-Jun (H-79) anti IκBα (C-21) antibodies were from Santa Cruz Biotechnologies; anti-H3K4me3 (ab8580) and anti-c-fos (ab7963) antibodies were from Abcam; anti-p-c-Jun (Ser73) (catalog # 9164) was from Cell Signaling.”

direct quote[Methods, Reagents]

“RAW264.7 mouse macrophage cell line was from ATCC (TIB-71).”

direct quote[Methods, Mice]

“Sample size estimate was based on power analyses performed using SAS 9.3 TS X64_7PRO platform.”

WARNReviewer 3· 85% conf

Antibodies, reagents, cell lines, and software are identified with some vendor and catalog numbers, but key resources such as RRIDs, lot numbers, and validation data for antibodies are missing. The miR-34a mimic and chitosan nanoparticles are partially described.

Evidence

direct quote[Methods (Reagents section)]

“RAW264.7 mouse macrophage cell line was from ATCC (TIB-71).”

direct quote[Methods (Bone Analyses, Statistical Analyses)]

“Micro-Computed Tomography (μCT) was performed using a Scanco μCT-35 instrument (SCANCO Medical)... Sample size estimate was based on power analyses performed using SAS 9.3 TS X64_7PRO platform.”

Tests are named (t-test, ANOVA, Mann-Whitney) but post-hoc tests are not stated. No assumption checks. Only p-value thresholds reported. No confidence intervals for effect sizes. Data are shown as bar graphs with error bars (SD/SEM) but no individual data points. SAS version is given.

Sub-criteria

tests namedINADEQUATE

exact p valuesINADEQUATE

effect sizes ciINADEQUATE

data presentationINADEQUATE

software identifiedINADEQUATE

assumptions verifiedNOT REPORTED

mathematical plausibilityN/A

Per reviewer (3) — the votes behind the verdict

FAILReviewer 1· 85% conf

Statistical reporting is incomplete: exact p-values are not given, assumptions not verified, effect sizes without confidence intervals, and data presentation uses bar graphs without individual points. Only the test names and software are partially identified.

Evidence

direct quote[Statistical Analyses]

“The p values were designated as: *, p<0.05; **, p<0.01; ***, p<0.005; ****, p<0.001; n.s. non-significant (p>0.05).”

absence[Methods]

Assumptions not verified.

direct quote[Figure legends]

“Error bars, SD.”

PASSReviewer 2· 85% conf

Statistical methods are generally appropriate and reported, with tests named and software identified, but there are some omissions: exact p-values are often reported as ranges, and assumptions are not verified.

Evidence

direct quote[Statistical Analyses, Methods]

“All statistical analyses were performed with Student's t-Test and represented as mean ± standard deviation (SD) unless noted otherwise.”

direct quote[Statistical Analyses, Methods]

“The p values were designated as: *, p<0.05; **, p<0.01; ***, p<0.005; ****, p<0.001; n.s. non-significant (p>0.05).”

absence[Statistical Analyses and results throughout]

Effect sizes and confidence intervals are not reported.

WARNReviewer 3· 80% conf

Statistical tests are named for most analyses (t-test, ANOVA, Mann-Whitney), but exact p-values are reported inconsistently (often as thresholds not exact values), effect sizes/CI are largely absent, and data presentation uses bar graphs with SD/SEM defined inconsistently. No arithmetic implausibility detected.

Evidence

direct quote[Methods (Statistical Analyses) and Figure legends (e.g., 3l-s)]

“All statistical analyses were performed with Student's t-Test and represented as mean ± standard deviation (SD) unless noted otherwise... l-s, Statistical analyses were performed with Mann Whitney Test... a,c,d,f,h-k,l,n,o,p p<0.05 by ANOVA.”

direct quote[Figure 3 and other figures]

“p<0.05 by ANOVA.”

direct quote[Figure legends]

“Error bars, SD. (in Figure 1, 2, 4); n=8, mean ± s.e. (in Extended Data 1e, 4c).”

absence[Results and Figures]

Effect sizes and confidence intervals are not reported.

There is no data availability section. No mention of deposition to GEO, SRA, or other repositories. No custom code identified or shared.

Sub-criteria

code sharingNOT REPORTED

accession numbersNOT REPORTED

repository depositNOT REPORTED

data availability statementNOT REPORTED

Per reviewer (3) — the votes behind the verdict

FAILReviewer 1· 95% conf

No data availability statement, repository deposit, accession numbers, or code sharing are provided. The paper does not mention any avenue for accessing data or code.

Evidence

absence[End of paper]

No data availability statement.

absence[All sections]

No repository deposit.

absence[All sections]

No code sharing.

WARNReviewer 2· 90% conf

Data availability is limited to 'on request', and no code is shared, which falls short of current standards for reproducibility.

Evidence

absence[Methods and results]

No data availability statement is present; data available on request is implied but not stated.

absence[Methods]

Custom code is not shared.

absence[Methods and results]

Data are not deposited in a repository.

FAILReviewer 3· 95% conf

No data availability statement, no repository deposit, no accession numbers, and no code sharing are reported. The paper does not provide any access to raw data or custom code.

Evidence

absence[Entire paper]

No data availability statement is present.

absence[Methods, Data Availability]

No repository deposit or accession numbers are reported.

absence[Methods]

No code sharing or mention of custom code.

Methods are described in extensive detail, covering mouse genetics, cell culture, histology, and statistical analyses. Limitations of the study are discussed (e.g., miR-34a may target other genes, but Tgif2 is key). Conclusions are proportional to the evidence, with appropriate caveats (e.g., 'suggest that' rather than definitive claims). Funding sources and conflicts of interest are reported. All outcomes mentioned in the methods appear to be reported. Reporting guidelines (ARRIVE) are not mentioned. Trial registration is not applicable as this is not a clinical trial.

Sub-criteria

funding coiADEQUATE

trial registrationN/A

reporting guidelineNOT REPORTED

methods completenessADEQUATE

all outcomes reportedADEQUATE

limitations discussedADEQUATE

conclusions proportionalADEQUATE

Per reviewer (3) — the votes behind the verdict

PASSReviewer 1· 85% conf

Methods are detailed, all outcomes are reported, conclusions are proportional, and funding/COI are stated. However, no reporting guideline (ARRIVE) is referenced, and limitations are only briefly discussed.

Evidence

absence[Methods]

No reporting guideline referenced.

direct quote[Acknowledgements]

“The authors declare that they have no financial conflict of interest.”

paraphrase[Results, Extended Data Figure 9]

“Osteoblastic miR-34a over-expression is not sufficient to inhibit osteoporosis or bone metastases.”

PASSReviewer 2· 90% conf

Methods are detailed, limitations are discussed, conclusions are proportional, and funding/COI are declared, though trial registration and reporting guidelines are not applicable.

Evidence

direct quote[Methods section, entire]

“The Methods section spans multiple paragraphs covering mice generation, cell culture, luciferase assays, etc.”

direct quote[Discussion, paragraph 3]

“Although miR-34a may also target genes other than Tgif2 in osteoclasts, our genetic rescue ex vivo and in vivo show that Tgif2 is the key miR-34a target, suggesting that other genes are likely secondary or functionally irrelevant to osteoclastogenesis.”

direct quote[Acknowledgements and Author contributions]

“The authors declare that they have no financial conflict of interest.”

WARNReviewer 3· 85% conf

Methods are detailed and complete, limitations are discussed, conclusions are proportional to evidence, and funding/COI are reported. However, the paper is a preclinical study that does not reference an appropriate reporting guideline (ARRIVE), and all outcomes do not appear to be pre-specified or reported against a pre-analysis plan.

Evidence

direct quote[Methods section (several sub-sections)]

“To generate cre-flox controlled conditional miR-34a transgenic mice (CAG34a), a 431 base-pair genomic sequence containing 168 base pairs 5′ and 161 base pairs 3′ of the pre-miR-34a sequence was inserted into the CAG-Z-EGFP vector.”

absence[Methods]

No reporting guideline is referenced (e.g., ARRIVE for animal studies).

direct quote[Discussion (final two paragraphs)]

direct quote[Acknowledgments]

“This work was in part supported by CPRIT (RP130145, YW; R1008, JM)... The authors declare that they have no financial conflict of interest.”

13 copyedit issues flagged: mostly typo, consistency, punctuation.

Checked 30 references: 30 verified.

Αlpha¹ Rigor Action ItemsFinal pass before journal submission

The published paper has moderate robustness but significant reporting gaps that an informed reader should weigh. The most consequential gaps are the complete absence of a data availability statement and code sharing, which severely limits reproducibility. The missing details on randomization, blinding, and replicate distinction also weaken confidence in the study design. A correction or erratum addressing data availability and key resource identification would strengthen the paper.

1.
HIGHdata code
Add a data availability statement in the Methods section, depositing raw μCT scans, imaging data, and RNA-seq data in a public repository (e.g., GEO, Figshare) with a DOI.
Without a data availability statement and repository deposit, the study's reproducibility is severely limited, and this is a critical gap for a data-driven paper.
2.
HIGHdata code
Deposit custom code used for analysis (e.g., μCT parameter scripts, statistical modeling) in a public repository (e.g., GitHub) and cite it in the paper.
Code sharing is essential for reproducibility, and its absence is a critical gap.
3.
HIGHrigor
Specify the randomization method (e.g., computer-generated random number table) and the randomization unit (e.g., individual mouse, cage) in the Statistical Analyses section.
The current statement 'Samples were randomly allocated' is insufficient for reproducibility and is a common reviewer concern.
4.
HIGHrigor
Describe blinding at all stages (allocation, treatment, outcome assessment) or justify why not feasible in the Methods section.
Blinding is only stated for analysis, not for allocation or treatment, which can introduce bias.
5.
HIGHrigor
Distinguish biological vs technical replicates in figure legends and methods.
Without this distinction, the statistical analysis and interpretation of variability are unclear.
6.
HIGHrigor
Provide RRIDs for all antibodies and list lot numbers or validation data in the Methods section.
Antibodies are identified with catalog numbers but lack RRIDs and validation data, which are essential for reproducibility.
7.
HIGHrigor
Authenticate the RAW264.7 cell line (e.g., STR profiling) and report the result, and report mycoplasma testing for all cell cultures.
Cell line authentication and mycoplasma testing are standard requirements for cell-based studies.
8.
HIGHstatistics
Report exact p-values (e.g., p=0.032) instead of thresholds (e.g., p<0.05) in figures and text, or provide a supplementary table with exact values.
Threshold-only p-values are less informative and do not allow readers to assess the strength of evidence.
9.
HIGHstatistics
Report effect sizes (e.g., Cohen's d) and 95% confidence intervals for key comparisons (e.g., BV/TV differences between WT and KO).
Effect sizes and confidence intervals provide a more complete picture of the results than p-values alone.
10.
HIGHstatistics
Verify and report statistical test assumptions (normality, homogeneity of variance) and actions taken if violated in the Statistical Analyses section.
Without assumption checks, the validity of the statistical tests is uncertain.
11.
MEDIUMreporting
Reference the ARRIVE guidelines for animal research in the Methods section.
Adherence to reporting guidelines improves completeness and transparency for animal studies.
12.
MEDIUMreporting
Add a dedicated limitations section discussing potential biases, confounding, and generalizability.
While limitations are briefly mentioned in the Discussion, a dedicated section would improve transparency.
13.
MEDIUMreporting
Add a statement on housing conditions for mice (e.g., light/dark cycle, temperature, humidity, cage type, enrichment) in the Methods section.
Housing conditions are a key biological variable that can affect experimental outcomes.
14.
MEDIUMreporting
Standardize error bar definitions (SD vs SEM) across all figure legends and justify the choice.
Inconsistent error bar definitions can confuse readers and affect interpretation of variability.
15.
LOWcopyedit
Fix typo 'osteoclastogenessi' to 'osteoclastogenesis' in Methods, Identification of miR-34a Targets.
Typographical errors can undermine the professionalism of the manuscript.
16.
LOWcopyedit
Remove stray commentary lines from References section (e.g., 'See Nature. 2020 Jun;582(7810):134 PMC7336956 . See commentary ...').
These lines appear to be editorial notes and do not belong in the manuscript.
17.
LOWcopyedit
Standardize punctuation in figure legends (e.g., 'Error bars, SD.' to 'Error bars represent mean ± SD').
Consistent formatting improves readability.

Full text (50,157 chars)Analyzed: Jul 6, 2026

Αlpha¹ assesses methodological rigor, not the importance of the findings. See how we calibrate →