Home

MetaAnaly

The Platform-Independent Computational Tool For Meta-analysis In The Paradigm of New Statistics

By W. J. Zhang



The user manual guide and suggested citation of this page:
Zhang W. J. 2024. MetaAnaly: The platform-independent computational tool for meta-analysis in the paradigm of new statistics. Network Biology, 14(2): 187-214
Also, click here to download the corresponding offline tool.




Fixed-effects Model
General Methods

Choose a weighting method:
Inverse-Variance (I-V) Method Inverse-Variance - Sample Size Hybrid (IVSS) Method

Number of studies (k):

Choose the p value for confidence intervals in forest plot:
p=0.01 p=0.001 p=0.05

Layout scale of forest plot (80, 60, 40, etc.):

Data of meta-analysis:

Reset and enter or copy k lines of space delimited data into this area.
The 1st column are study IDs; the 2nd column are effect sizes θi of k studies.
The 3rd column are standard errors Sθi of effect sizes of k studies.
The 4th column are sample sizes ni of of k studies.
If sample sizes are not available, the 4th column should be set as 0's.



Results of meta-analysis for fixed-effects model:



I-V method and IVSS method can be used to pool almost all types of effect measures, including continuous measures, OR, RR, HR, etc. In the forest plot, you may change the study names by clicking and editing the study ID cells in the "Study" column.

User manual guide:
Zhang W. J. 2024. MetaAnaly: The platform-independent computational tool for meta-analysis in the paradigm of new statistics. Network Biology, 14(2): 187-214

   Back to Top   


Fixed-effects Model
Ratio/Risk Methods

Choose an effect measure:
Odds Ratio (OR) Relative Risk (RR) Risk Difference (RD) Hazard Ratio (HR)

Choose a weighting method:
Mantel-Haenszel (M-H) Method Peto Method

Number of studies (k):

Data of meta-analysis:

Reset and enter or copy k lines of space delimited data into this area.
The 1st column are study IDs; the 2nd column are ORi, RRi, or RDi of k studies.
The 3rd column are numbers of cases, ai's, with outcome events in group A for k studies.
The 4th column are numbers of cases, bi's, without outcome events in group A for k studies.
The 5th column are numbers of cases, ci's, with outcome events in group B for k studies.
The 6th column are numbers of cases, di's, without outcome events in group B for k studies.



Results of meta-analysis for fixed-effects model:


Recommended usage for the two weighted averaging methods include: (1) When pooling RR and RD in binary variables, the M-H method can be used. (2) When pooling OR in binary variables, the M-H method and the Peto method can be used. (3) When pooling HR values in survival data, the Peto method can be used.

User manual guide:
Zhang W. J. 2024. MetaAnaly: The platform-independent computational tool for meta-analysis in the paradigm of new statistics. Network Biology, 14(2): 187-214

   Back to Top   


Random-effects Model

Choose a method:
DerSimonian-Laird (DL) Estimator
Hunter-Schmidt (HS) Estimator
Maximum-likelihood (ML) Estimator
Restricted Maximum-likelihood (REML) Estimator
Modified Knapp-Hartung (mKH) Estimator
Paule-Mandel (PM) Estimator
All Estimators
Averaged Estimator

Number of studies (k):

Choose the p value for confidence intervals in forest plot:
p=0.01 p=0.001 p=0.05

Layout scale of forest plot (80, 60, 40, etc.):

Data of meta-analysis:

Reset and enter or copy k lines of space delimited data into this area.
The 1st column are study IDs; the 2nd column are effect sizes θi of k studies.
The 3rd column are standard errors Sθi of effect sizes of k studies.
The 4th column are sample sizes ni of of k studies.
If sample sizes are not available, the 4th column should be set as 0's.



Results of meta-analysis for random-effects model:



Currently the averaged estimator is based on the averaging estimation of the six estimators above. More estimators are expected to be used in the future revisions. For averaged estimator, I recommend to use the results that all estimators are assumed to be homogeneous. If you are sure that for some data these estimators will yield significantly different results, you can use the results that all estimators are assumed to be heterogeneous. In the forest plot, you may change the study names by clicking and editing the study ID cells in the "Study" column. The forest plot will not be available if "All Estimators" is chosen.

User manual guide:
Zhang W. J. 2024. MetaAnaly: The platform-independent computational tool for meta-analysis in the paradigm of new statistics. Network Biology, 14(2): 187-214

   Back to Top   



Use one of the following methods:
Q test I2 I Test I2 II Test CH Test Cohen’s d Test Hedges’ g Test

α value for χ2 test in Q test:
0.1 0.05

Number of studies (k):

Data of meta-analysis:

Reset and enter or copy k lines of space delimited data into this area.
The 1st column are study IDs; the 2nd column are effect sizes θi of k studies.
The 3rd column are standard errors Sθi of effect sizes of k studies.
The 4th column are sample sizes ni of of k studies.
If sample sizes are not available, the 4th column should be set as 0's.



Results of heterogeneity testing:


Dr. H. N. Huang speculated that there may be something wrong with the I2 I Test, which seems to be biased and can be further studied.

User manual guide:
Zhang W. J. 2024. MetaAnaly: The platform-independent computational tool for meta-analysis in the paradigm of new statistics. Network Biology, 14(2): 187-214

   Back to Top   



Subgroups Analysis

Subgroup analysis can be conducted using ANOVA online tool.

User manual guide:
Subgroup Analysis: Zhang W. J. 2024. MetaAnaly: The platform-independent computational tool for meta-analysis in the paradigm of new statistics. Network Biology, 14(2): 187-214
ANOVA: Zhang W. J., Qi Y. H. 2024. ANOVA-nSTAT: ANOVA methodology and computational tool in the paradigm of new statistics. Computational Ecology and Software, 14(1): 48-67.

   Back to Top   


Meta-Stepwise Regression

Number of suspected heterogeneity factors (n):

Number of studies (k):

Margin F value for inclusion and exclusion of suspected heterogeneity factors:
0.3 0.9 0.8 0.7 0.6 0.5 0.4 0.2 0.1

Data of meta-stepwise regression:

Reset and enter or copy m lines of space delimited data into this area.
The 1st column are study IDs, the following n columns are suspected heterogeneity factors.
The last column (the column n+2) are effect sizes of m studies.



Results of meta-stepwise regression:


User manual guide:
Meta-regression: Zhang W. J. 2024. MetaAnaly: The platform-independent computational tool for meta-analysis in the paradigm of new statistics. Network Biology, 14(2): 187-214
Stepwise Regression: Qi Y. H., Liu G. H., Zhang W. J. 2016. A Matlab program for stepwise regression. Network Pharmacology, 1(1): 36-40.

   Back to Top   



Cohen’s d Test for Post Meta-analysis

Effect size of A:

Effect size of B:

Standard error of effect size of A:

Standard error of effect size of B:



Results of Cohen’s d test for difference significance between two effect sizes:


User manual guide:
Difference Significance Between Two Effect Sizes: Zhang W. J. 2024. MetaAnaly: The platform-independent computational tool for meta-analysis in the paradigm of new statistics. Network Biology, 14(2): 187-214
Calculation of Effect Sizes: Zhang WJ. 2023. A desktop calculator for effect sizes: Towards the new statistics. Computational Ecology and Software, 13(4): 136-181.

   Back to Top   


Hedges’ g Test for Post Meta-analysis

Effect size of A:

Effect size of B:

Standard error of effect size of A:

Standard error of effect size of B:

Sample size of A:

Sample size of B:



Results of Hedges’ d test for difference significance between two effect sizes:


User manual guide:
Difference Significance Between Two Effect Sizes of Same Quantity: Zhang W. J. 2024. MetaAnaly: The platform-independent computational tool for meta-analysis in the paradigm of new statistics. Network Biology, 14(2): 187-214
Calculation of Effect Sizes: Zhang WJ. 2023. A desktop calculator for effect sizes: Towards the new statistics. Computational Ecology and Software, 13(4): 136-181.

   Back to Top   


Copyright © 2024 - W. J. Zhang (E-mail: wjzhang@iaees.org)
International Academy of Ecology and Environmental Sciences. E-mail: office@iaees.org
Copyright © 2009-2024. International Academy of Ecology and Environmental Sciences. All rights reserved.
Web administrator: office@iaees.org, website@iaees.org;


Translate page to: