Written and maintained by CASRAI Editorial Board
Last updated
Elisabeth Bik is a microbiologist whose published work on research integrity centres on screening biomedical papers for inappropriately duplicated or manipulated images. This guide covers that work and its methods: the 2016 mBio study she co-authored, the three categories of image duplication it defined, what a visual screen can and cannot show, and how a reader or editor can act on an image concern. It is written for editors, research integrity officers and researchers who need to understand what the study found and how to use it, not as a personal profile.
Who is Elisabeth Bik, in terms of the work
Science Integrity Digest, the blog she publishes through Harbers-Bik LLC, describes itself as a blog about science integrity written by microbiologist Elisabeth Bik. Its stated scope is her analyses of papers (mainly image problems), searches of the biomedical literature for inappropriately duplicated or manipulated photographic images and for plagiarized text, and guides for readers. The site says most of this work is unpaid and supported through Patreon (read 7 November 2026).
In the research-integrity literature her name is most closely tied to one paper, described next.
The 2016 mBio study: method and findings
Short answer: Bik, Casadevall and Fang screened 20,621 papers and found problematic figures in 782 of them (3.8%, roughly 1 in 25).
The paper is Bik EM, Casadevall A, Fang FC, “The Prevalence of Inappropriate Image Duplication in Biomedical Research Publications,” mBio 7(3):e00809-16, published 7 June 2016 (open copy on PubMed Central, read 7 November 2026).
- Sample: 20,621 papers containing the term “Western blot”, from 40 journals and 14 publishers, published 1995 to 2014. PLoS One supplied 8,138 papers (39.5%) of the 2013 to 2014 portion.
- Result: 782 papers (3.8%) contained problematic figures.
- Trend: the rate rose sharply from 2003 and stayed near or above 4% afterwards.
- By journal: from 0.3% (Journal of Cell Biology) to 12.4% (International Journal of Oncology), inversely correlated with journal impact factor.
- Repeat authors: about 39% of authors of problematic papers had another paper with duplication.
The three categories of image duplication
The study sorted the 782 problematic papers into three categories. The counts below are the study’s own.
| Category | What it describes | Papers |
|---|---|---|
| I. Simple duplication | The same image appears again in the same paper, unaltered. | 230 (29.4%) |
| II. Duplication with repositioning | A duplicated image that was shifted, rotated or reversed. | 356 (45.5%) |
| III. Duplication with alteration | Duplicated lanes, bands or cells, including stamping or patching. | 196 (25.1%) |
For the shared vocabulary, see the CASRAI entries for image duplication and image manipulation.
What visual screening can and cannot show
The authors state the limits of the design themselves:
- It examined images only, not other data.
- It looked for duplication within a single paper, not between different papers.
- It relied on visual inspection, so the number of missed cases is unknown and the 3.8% figure is likely an undercount.
- It cannot determine intent. A duplicated image can come from an honest error or from deliberate manipulation, and only an institutional or journal inquiry can separate them.
- The sample was dominated by PLoS One and by papers found through the “Western blot” search term, so it is not a random sample of all biomedical publishing.
The practical reading: a flagged image is a reason to ask the authors and the journal for the original data, not a finding of misconduct.
Science Integrity Digest and how to raise an image concern
Science Integrity Digest publishes reader guides on checking for plagiarism, posting concerns on PubPeer and reporting suspected misconduct to editors. Two routes follow from that:
- Post the specific concern publicly on PubPeer, naming the figure and panels involved.
- Report it to the journal’s editor, who can follow the Committee on Publication Ethics (COPE) flowchart “Inappropriate image manipulation in a published article” (version 2, 20 May 2024). The flowchart is a procedural guide for editors, not a forensic method.
The International Committee of Medical Journal Editors names deceptive image manipulation as a form of misconduct, and the long-standing editorial rule from Rossner and Yamada (Journal of Cell Biology 166(1):11-15, 2004) is that any adjustment must apply to every pixel of the image.
Strengths and limits, and related tools
Human visual screening of the kind used in the 2016 study identified three distinct duplication patterns across thousands of papers, but it is slow and its misses cannot be counted. Software now automates part of this work. CASRAI’s guides to Proofig and Imagetwin cover two such tools, and responsible use of AI image detectors covers how to treat their output.
Frequently asked questions
What did Elisabeth Bik’s 2016 study find?
Of 20,621 papers screened, 782 (3.8%) had problematic figures, and the rate rose from 2003 onward (Bik, Casadevall and Fang, mBio, 2016).
What is Science Integrity Digest?
A blog on science integrity written by Elisabeth Bik and published through Harbers-Bik LLC, covering image problems, plagiarized text and reader guides.
Does a duplicated image prove misconduct?
No. The study’s authors note that intent cannot be determined from the images alone.
How do I report a suspected image problem?
Post a specific, evidence-based concern on PubPeer and report it to the journal’s editor, who can use the COPE image-manipulation flowchart.
Sources
- Bik EM, Casadevall A, Fang FC. mBio 7(3):e00809-16 (2016): PMC4941872, read 7 November 2026.
- Science Integrity Digest, About, read 7 November 2026.
- COPE, flowchart “Inappropriate image manipulation in a published article,” v2, 20 May 2024.
