5. Meta-analysis - Meta-regression with continuous covariates and categorical factors

Рет қаралды 2,263

Күн бұрын

This video in this series on meta-analysis explains why and how to perform a meta-regression analysis to explain for the heterogeneity in effect sizes in the studies which are not accounted for even on doing a random-effects restricted maximum likelihood analysis. Both covariates and factors are used to perform a multivariate meta-regression analysis.
This series is all about the basics of meta-analysis, when to use the fixed-effects model versus the random-effects model, the omnibus test of model coefficients, the Q-test for heterogeneity, tau squared statistics, I squared statistics, forest plots, funnel plots, rank test for publication bias and funnel plot asymmetry, Egger's regression test for publication bias and funnel plot asymmetry, Fail-Safe N approach, trim-fill analysis, meta-regression using continuous covariates and categorical factors, how forest and funnel plots change with meta-regression and how to interpret them, and exhaustive reporting for the meta-analysis results.
The Campbell collaboration website - www.campbellco...
The analysis shown here is done in the statistical software package - JASP
The JASP website: jasp-stats.org/
The JASP download page: jasp-stats.org...
How to download and install JASP: • 1. How to download + i...
How to convert and upload a Microsoft Excel file to JASP: • 3. How to upload a dat...
Types of data and variables: • 4. Types of data and v...
Which statistical test to use: • 5. Which statistical t...
Null and alternative hypotheses: • 6. What are null and a...
p-value and statistical significance: • 7. The p-value - trans...
Writing up the final report: • 8. Final reporting - w...
Bayesian vs. Frequentist analysis: • Introduction to Bayesi...
When should you use Bayesian analysis: • When should you use Ba...
This tutorial shows how statistical analysis can be done free of cost, by anyone, irrespective of their mathematical aptitude. These tutorials are made especially for medical students and residents or any researcher in health-care, who need data quickly analyzed, with tables, graphs, and plots conveying maximum information, and also reporting those results in a standard manner, so that writing the 'materials and methods' and 'results' section of the manuscript becomes a cake-walk.
Statistics, especially biostatistics, is an enigma all throughout undergraduate days for medical students. Applying a few formulae which didn't even make sense during a 15 minute biostatistics exam was all that was expected.
When it comes to research however, students often employ the services of a statistician to help analyze their data.
Seldom do students do statistical analyses on their own.
The following video series will help medical students to analyze their own data.
Our tool of choice is going to be JASP software, which is an excellent and free, open-source alternative to the gold-standard SPSS software, which needs users to pay to use the software - a luxury students cannot afford!
The following series is going to focus on how to use the software to do statistical analyses, quite unlike what is taught in most medical schools. I hope that this series will help medical students to take up research early on in their undergraduate days, and also analyze their own data by themselves, completely independently. I also hope that this will change the perspective of biostatistics that medical students have, which is inculcated by the syllabus studied during their MBBS days.