Observational Subgroup Meta-Analysis Support
Budget: £20 – £250 GBP
This project centres on a systematic review and meta-analysis that will synthesise results exclusively from observational studies. The chief objective is to quantify and clearly present subgroup differences; every analytical decision should therefore keep comparative subgroup interpretation front-and-centre.
Scope of work
• Confirm or refine my existing search strategy, ensuring consistent inclusion/exclusion criteria for observational data.
• Build a clean, analysable dataset (effect estimates and variances) ready for R, Stata, or RevMan.
• Select the appropriate fixed- or random-effects model, evaluate heterogeneity (I², τ²), explore publication bias, and run sensitivity checks.
• Generate forest plots, subgroup-specific pooled estimates, and—where useful—meta-regression outputs to illuminate between-group variability.
• Deliver a concise statistical report that can drop directly into a manuscript, including methods text, tables, and graphics.
Acceptance criteria
1. Reproducible analysis script (R Markdown or do-file).
2. Annotated output covering each subgroup comparison, heterogeneity metrics, and model choice rationale.
3. Publication-ready figures (300 dpi PNG & vector).
4. Brief memo summarising key findings and any limitations encountered.
When you respond, please attach examples of past meta-analyses—particularly those that handled subgroup work in observational contexts—so I can quickly gauge fit.
Scope of work
• Confirm or refine my existing search strategy, ensuring consistent inclusion/exclusion criteria for observational data.
• Build a clean, analysable dataset (effect estimates and variances) ready for R, Stata, or RevMan.
• Select the appropriate fixed- or random-effects model, evaluate heterogeneity (I², τ²), explore publication bias, and run sensitivity checks.
• Generate forest plots, subgroup-specific pooled estimates, and—where useful—meta-regression outputs to illuminate between-group variability.
• Deliver a concise statistical report that can drop directly into a manuscript, including methods text, tables, and graphics.
Acceptance criteria
1. Reproducible analysis script (R Markdown or do-file).
2. Annotated output covering each subgroup comparison, heterogeneity metrics, and model choice rationale.
3. Publication-ready figures (300 dpi PNG & vector).
4. Brief memo summarising key findings and any limitations encountered.
When you respond, please attach examples of past meta-analyses—particularly those that handled subgroup work in observational contexts—so I can quickly gauge fit.