ScribeLab Writer

Meta-Analysis Statistical Analysis & Systematic Review Source Screening

ScribeLabWriter provides specialist statistical analysis and source screening support for researchers conducting systematic reviews and meta-analyses. If you have your studies identified and your data extracted but need expert support with the quantitative synthesis — or if you need rigorous, documented screening of a large body of retrieved literature — our team of PhD-qualified statisticians and research methodologists delivers precise, publication-ready output aligned with your research question, your data, and the reporting standards of your target journal. This service can be engaged as a standalone component or in combination with our full systematic review writing service.

Meta-Analysis Statistical Analysis

  • Effect Size Calculation and Pooled Estimates We calculate appropriate effect measures for your data type — risk ratios (RR), odds ratios (OR), mean differences (MD), standardised mean differences (SMD), or other measures as required — and pool estimates across included studies using fixed-effects or random-effects models as appropriate to your data. We select the pooling model transparently and justify the choice in the methods section.

  • Forest Plot Generation We produce publication-ready forest plots showing individual study estimates with confidence intervals, study weights, and the pooled effect estimate with its 95% confidence interval. Forest plots are formatted to the specifications of your target journal and provided as high-resolution figures suitable for submission.

  • Heterogeneity Assessment We assess and report between-study heterogeneity using the I² statistic, Cochran's Q test, and tau² (between-study variance). Where heterogeneity is substantial, we explore potential sources and report findings transparently in accordance with PRISMA 2020 and Cochrane reporting standards.

  • Subgroup Analysis Where your research question or reviewer requirements call for it, we conduct pre-specified or exploratory subgroup analyses to investigate whether effect sizes vary meaningfully across participant groups, interventions, settings, study designs, or other variables of interest. All subgroup analyses are reported with appropriate caution regarding multiple testing.

  • Sensitivity Analysis We conduct sensitivity analyses to assess the robustness of pooled estimates — for example, by excluding high-risk-of-bias studies, using alternative effect measures, or applying different inclusion criteria. Sensitivity analysis is essential for demonstrating the reliability of your conclusions to peer reviewers and editors.

  • Publication Bias Assessment We assess potential publication bias using funnel plots, Egger's test, and Begg's test where the number of included studies is sufficient. Findings are reported transparently with appropriate interpretation of asymmetry and its possible causes.

  • GRADE Summary of Findings Tables We apply the GRADE framework to assess the certainty of evidence for each outcome in your review and produce Summary of Findings (SoF) tables in the format required by your target journal. GRADE assessment is required by Cochrane reviews and increasingly expected by high-impact journals publishing systematic evidence syntheses.

  • Statistical Software We work in R (meta, metafor packages), RevMan (Cochrane Review Manager), Stata, and SPSS depending on your project requirements and target journal preferences. All code and output files are provided on request.

Systematic Review Source Screening

The screening process — applying pre-specified inclusion and exclusion criteria to retrieved records at the title/abstract stage and then at full-text stage — is one of the most labour-intensive components of a systematic review. It requires methodological consistency, documented decision-making, and a reproducible audit trail that can withstand scrutiny from peer reviewers and editors. We manage systematic source screening using Rayyan and Covidence, the two most widely used screening platforms in systematic review methodology.

  • Title and Abstract Screening We screen the full set of records retrieved by your database searches against your inclusion and exclusion criteria at the title and abstract level, documenting include, exclude, and uncertain decisions for every record. Uncertain records are resolved through discussion and the rationale recorded.

  • Full-Text Eligibility Assessment Records passing title/abstract screening are retrieved and assessed against your inclusion and exclusion criteria at full-text level. All exclusion decisions at this stage are documented with the reason for exclusion, which is required for the PRISMA flow diagram.

  • PRISMA Flow Diagram We generate a complete PRISMA 2020 flow diagram documenting the number of records identified through each database search and additional sources, the number of duplicate records removed, the number screened at each stage, and the final number of studies included in the review.

  • Screening Documentation We provide a full screening log documenting all decisions made during the screening process, suitable for submission as supplementary material if required by the journal.

What This Service Does Not Include

This service covers statistical analysis and source screening as standalone components. It does not include writing the systematic review manuscript, developing the search strategy, or conducting the database searches. If you need support with writing the complete manuscript, please see our Full Writing Service. If you need manuscript revision based on peer reviewer comments, please see our Manuscript Revision Service.

Frequently Asked Questions

We conduct meta-analysis using R (meta, metafor packages), RevMan, Stata, and SPSS depending on the requirements of your project and target journal. We select the most appropriate software for your study types, effect measure, and the specific analyses required, and provide all output in the format required for journal submission.

A forest plot is a graphical display of the results of individual studies included in a meta-analysis alongside the pooled effect estimate. Each study is represented as a horizontal line (confidence interval) and a square (point estimate weighted by study size), with the pooled effect shown as a diamond at the bottom. Forest plots are the standard method for presenting meta-analysis results and are required by virtually all journals publishing quantitative systematic reviews. We produce publication-ready forest plots formatted to your target journal's specifications.

Heterogeneity in meta-analysis refers to variability in effect sizes across included studies beyond what would be expected by chance alone. We assess heterogeneity using the I² statistic, Cochran's Q test, and tau². When substantial heterogeneity is present (I² > 50%), we explore potential sources through subgroup analysis and meta-regression, and report findings transparently in accordance with PRISMA 2020 and Cochrane reporting standards.

GRADE — Grading of Recommendations Assessment, Development and Evaluation — is the internationally recognised framework for assessing the certainty of evidence synthesised in systematic reviews. It rates the quality of evidence as high, moderate, low, or very low based on risk of bias, inconsistency, indirectness, imprecision, and publication bias. GRADE assessments are presented in a Summary of Findings table. We apply the GRADE framework and produce SoF tables as part of our statistical analysis service.

We conduct systematic source screening using Rayyan and Covidence — the two most widely used tools for managing title, abstract, and full-text screening in systematic reviews. Both platforms allow for transparent, reproducible, and auditable screening decisions documented in the PRISMA flow diagram.

Related Reading

Need statistical analysis or screening support?

Describe your project and requirements and we will respond with a detailed quote within 6 business hours.

Request a Quote

GET STARTED