What is Research Misconduct? Part 2: Falsification

This is Part 2 of a series of 3, which also includes Part 1: Plagiarism, and Part 3: Fabrication.

In Part 1 of this series, I showed some examples of plagiarism in scientific papers, which the Office of Research Integrity (ORI) considers one of the three forms of Research Misconduct. Here, we will look at the second type of misconduct, falsification. ORI defines falsification as follows:

Falsification is manipulating research materials, equipment, or processes, or changing or omitting data or results such that the research is not accurately represented in the research record.”

Office of Research Integrity: Definition of Research Misconduct

Note that “Falsification” and “Fabrication” are not always easy to distinguish. Fabrication is making up data, so reporting on experiments that never happened or patients that never existed. Falsification is different in that an experiment might have taken place, but that some measurements were altered. Although the definitions appear to be very different, the reuse of an image to represent two different experiments could be interpreted as either falsification (changing data) or as fabrication (making up data). For now, I will file these under falsification.

Here are some examples of research falsification; but again, note that several of these might also be interpreted as fabrication.

  • Removing an outlier from a series of measurements
  • Changing a measurement to make it look higher or lower
  • Adjusting the brightness of one particular gel band, but not the other bands in the same photo
  • Adding error bars to single measurements to suggest multiple measurements were performed
  • Re-using the same Western blot (either in the same orientation, or mirrored, rotated, or stretched) to represent a different experiment
  • Showing two overlapping microscopy images to represent 2 different experiments
  • Presenting data as if derived from cell line A while it was obtained from cell line B
  • Duplicating some cells on a photo to make it look like a treatment resulted in a higher cell density

Falsification cases in the news

Here are some well-documented cases where researchers had to retract papers because of data falsification. I will write more about these in future posts.

Image duplication suggestive of misconduct

One of the most often encountered- because most visible – problems with scientific papers are duplicated images. In a screen of 20,000 biomedical papers, my coauthors, Arturo Casadevall and Ferric Fang, and I found that 4% of them contained duplicated photos within the same paper (Bik et al. 2016 – The Prevalence of Inappropriate Image Duplication in Biomedical Research Publications – mBio 7(3):e00809-16. doi:10.1128/mBio.00809-16). Some of these duplicated panels were exact copies, and those could be honest errors. But others were mirrored/overlapping or even manipulated, which are suggestive of misconduct.

Here are some examples, all from papers that have been retracted. On the left you will see the original figure; on the right the figure (sometimes made a bit darker to enhance contrast) with colored boxes depicting areas that look very similar to each other.

In these photos, overlapping microscopy photos or similar-looking Western blots suggest that the same sample was used to represent multiple experiments.

Duplicated immunoblots.
Left: original image. Right: colored boxes show regions that look similar.
Overlapping microscopy photos.
Left: original image. Right: colored boxes show regions that look similar.
Duplicated Western blots.
Left: original image. Right: colored boxes show regions that look similar.

Image manipulation suggestive of misconduct

In the above example, I showed some images with potential regions of overlap. There are also cases in which the same feature is visible multiple times within the same photo. This is often called “cloning” by photographers. Or, in other cases, certain parts of 2 photo look very similar but other parts don’t, because certain parts appear to have been added or taken out. Here are some examples of possible photo manipulation (again, all from retracted papers):

The left part of this photo appears to contain “cloning” in which the same block is visible multiple times. In addition, sharp transitions and variable background patterns suggests this photo might be a composite.
Some areas of these photos look very similar, but other areas look different.
This actin blot appears to be contain just 2 different bands.

Falsification cases from ORI

The Office of Research Integrity regularly publishes case summaries of research misconduct, usually about research performed with USA federal grant money. Here are some cases of interest from the ORI website:

Further reading

5 thoughts on “What is Research Misconduct? Part 2: Falsification”

  1. Falsification is usually discussed in the context of altering data or images. What about the situation where researchers claim they have performed a systematic review of the literature on a topic but probably actually didn’t. They follow the PRISMA reporting guidelines which require among other things an electronic version of one of their searches. However, careful inspection suggests that the articles they include in their review were not selected by the search but were just hand-picked, and that the ‘search strategy’ they describe was actually reverse-engineered, not completely successfully, to identify the articles that they had already selected. Is that falsification if they didn’t really perform a systematic search but claim they did?

    Like

    1. Tough question, in part because I am not familiar with the guidelines on how to write these. It is falsification for sure if they handpicked favorable articles and not included specific articles, but it is going to be very hard to prove for a reviewer or reader.
      A lot of falsification is actually hard to prove unless you are sitting next to the researchers performing the experiments, or if you have detailed lab books to look in. Cases of falsification/fabrication that are caught are usually because authors left visible traces, such as duplicated cells in the same photo. It has me and others worried that many misconduct cases will remain unnoticed.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s