Kombucha Antioxidant Claims and Publication Bias Study

ISEF Category: Biochemistry

Subcategory: Other  ·  Difficulty: Advanced  ·  Setup: Home Setup  ·  Time: 1 to 2 Months

The Hook

A flashy result can shrink when you look at every study, not just the ones that got attention. That is a real risk with kombucha and other fermented-food antioxidant claims. Your project asks a sharp question: do the reported gains still hold after you correct for publication bias and small sample sizes? If the answer changes, you have found something that matters for how food science gets reported.

What Is It?

This project is a meta-analysis, which means you are combining results from published studies to see the bigger pattern. Instead of treating each paper as a final answer, you treat each one like a data point. That helps you ask whether the full literature supports the headline claim about antioxidant gains.

Think of the literature like a pile of puzzle pieces. A few bright, positive studies can make the picture look complete, even if weaker or negative studies never made it into print. Publication bias happens when studies with exciting results are more likely to be published, and small-sample effects happen when tiny studies swing to extreme results by chance. Your job is to check whether the antioxidant story stays strong after those distortions are corrected.

Why This Is a Good Topic

This topic works well because you can test a clear claim with public data, not a lab bench. You connect directly to a real-world question about fermented foods, health claims, and how trustworthy published results are. You can learn how to screen papers, extract numbers, compare effect sizes, and check whether a literature signal is stable or shaky.

Research Questions

• How does the reported antioxidant effect size change after correcting for publication bias in kombucha studies?
• What is the effect of small sample size on the reported antioxidant gains in fermented-food studies?
• Does the estimated antioxidant benefit differ between kombucha and other fermented foods?
• To what extent do assay type and outcome measure explain differences in reported effect sizes?
• Which studies have the largest influence on the pooled result after sensitivity analysis?
• How does publication year relate to the size of the reported antioxidant effect?
• To what extent do positive findings stay significant after outlier removal and bias correction?

Basic Materials

• Laptop with internet access and enough storage for PDFs and data files.
• Spreadsheet software such as Google Sheets or Excel for screening and data extraction.
• Reference manager such as Zotero for saving papers and notes.
• PDF reader with annotation tools for checking methods and results sections.
• Headphones or a quiet workspace for focused paper screening.
• External drive or cloud storage for backing up your review files.

Advanced Materials

• University library access for full-text journal articles behind paywalls.
• R with the metafor, dmetar, and tidyverse packages for meta-analysis and plotting.
• RStudio desktop or Posit Cloud for running and saving analysis scripts.
• Reference manager with shared-library support for team screening.
• Statistical consultation access for checking model choice and bias tests.
• High-resolution monitor or dual screens for paper screening and code review.

Software & Tools

• Zotero: Organizes papers, PDFs, and citation notes as you screen studies.
• PubMed: Finds primary studies and review articles on kombucha and antioxidant assays.
• R: Fits meta-analysis models, forest plots, funnel plots, and bias tests.
• RStudio: Gives you a clean workspace for coding, tables, and figures.
• Google Sheets: Tracks screening decisions, extracted variables, and study features.

Experiment Steps

1. Define the exact foods, outcomes, and study designs you will include.
2. Build a screening sheet that separates usable studies from reviews, duplicates, and papers without extractable data.
3. Choose one effect-size format and plan how you will convert different antioxidant measures into the same scale.
4. Decide which bias checks you will run, such as funnel plots, small-study tests, and sensitivity analyses.
5. Plan subgroup comparisons that could explain differences, such as food type, assay type, sample size, or publication year.
6. Set rules for removing weak studies so one outlier cannot drive the full conclusion.

Common Pitfalls

• Mixing different antioxidant assays without grouping them, which makes the pooled result hard to interpret.
• Counting multiple papers from the same experiment as separate studies, which gives one lab too much weight.
• Combining human, animal, and in vitro results in one pool, which hides the real pattern.
• Using only abstract numbers when the paper does not report enough detail for a stable effect size.
• Skipping small-study checks, which can make a loud result look more certain than it really is.

What Makes This Competitive

A class-level version just summarizes a few papers. A stronger version codes each study carefully, compares several bias-correction methods, and shows how stable the conclusion stays when you drop weak studies. You can push it further by separating assay types, food types, and sample sizes, then testing whether any subgroup still shows a real signal. That makes the project feel like evidence evaluation, not just literature summary.

Project Variations

• Compare kombucha papers with kefir, yogurt, or other fermented foods to see whether publication bias looks different across food types.
• Focus on one assay family, such as DPPH or FRAP, so the chemistry stays more consistent across studies.
• Test whether fermentation time, sugar source, or starter culture predicts reported antioxidant effect size across papers.

Learn More

• PubMed: Search for primary studies and reviews on kombucha, polyphenols, and antioxidant assays.
• NCBI Bookshelf: Read free chapters on meta-analysis, biostatistics, and evidence synthesis.
• PRISMA 2020 Statement: Use the reporting checklist for systematic reviews and meta-analyses.
• Cochrane Handbook for Systematic Reviews of Interventions: Read the methods chapters for study selection and bias checks.
• NIH Office of Dietary Supplements: Find background on flavonoids, polyphenols, and related bioactive compounds.
• metafor package documentation: Learn how to build forest plots, funnel plots, and publication-bias tests in R.