Falsification
Scientific Misconduct includes Fabrication, Falsification, and Plagiarism. A formal definition of Scientific Misconduct from the DHHS is still pending, however there is general consensus of the elements.
Definition: Falsification of data is the selective alteration of data collected in the conduct of scientific investigation or the "misrepresentation of uncertainty" during statistical analysis of the data. Falsification also includes the selective omission/deletion/suppression of conflicting data without scientific or statistical justification.
Examples of Falsification
- Alteration of data to render a modification of the variances in the data
- Falsification of dates and experimental procedures in the study notebook
- Misrepresenting the results from statistical analysis
- Misrepresenting the methods of an experiment such as the model (e.g., cell line) used to conduct the experiment
- The addition of false or misleading statements in the manuscript or published paper. For example the misrepresentation of "n".
- Falsification of research accomplishments by publishing the same research results in multiple papers (self plagiarism)
- Falsification of data in continuation application for research supported by PHS funds
- Misrepresentation of the materials or methods of a research study in a published paper
- Providing false statements about the extent of a research study in an abstract submitted for publication and oral presentation at a professional society meeting
- Falsification of telephone call attempts to collect data for a survey study such as in a federally funded program to determine risk factors for new mothers and babies
Examples of Falsification in Clinical Studies:
From Assessing Scientific Misconduct Allegations Involving Clinical Research
- Substituting one subject's record for that of another subject
- Falsely reporting to a data coordinating center that certain clinical trial staff, who were certified to perform the procedures on the subjects, had done so, when they had not
- Altering the dates and results from subjects' eligibility visits
- Altering the dates on patient screening logs and/or submitting the same log with altered dates on multiple occasions
- Failing to update the patients' status and representing data from prior contacts as being current
- Altering the results of particular tests on blood samples to show that the test accurately predicted a disease or relapse
- Backdating follow-up interviews to fit the time window determined by the study protocol
- Falsifying the times that blood samples were drawn from human subjects