Wikipedia defines: Fabrication, in the context of scientific inquiry and academic research, refers to the act of intentionally falsifying research results, such as reported in a journal article. Fabrication is considered a form of scientific misconduct, and is regarded as highly unethical. In some jurisdictions, fabrication may be illegal.
The word falsifying used above should not be confused with the legitimate and essential activity of finding and sharing evidence that contradicts a hypothesis (see falsifiability) but is used in the sense of deliberately presenting known false information as true with the intent to deceive. Neither should the concept be applied to a scientist or a group of scientists deceiving themselves; this behaviour is sometimes called pathological science.
Examples of activities which constitute fabrication include:
Outright synthesis of experimental data; reporting experiments which were never conducted. Sometimes referred to as “drylabbing”.
“Fudging”, “massaging”, or outright manufacture of experimental data.
Inappropriate, and statistically invalid, “culling” of experimental data, such as the intentional exclusion of experimental runs which contradict the hypothesis the scientist is trying to demonstrate, or excessive filtration of “noise” which suggests a correlation where none can be shown to exist.
Intentional portrayal of interdependent events as independent.
Ordering subordinates or research assistants to participate in any of the above.
In addition, some forms of (unintentional) academic incompetence or malpractice can be difficult to distinguish from intentional fabrication. Examples of this include the failure to account for measurement error, or the failure to adequately control experiments for the parameter(s) being measured.
‘“Fudging”, “massaging”, or outright manufacture of experimental data.
Inappropriate, and statistically invalid, “culling” of experimental data, such as the intentional exclusion of experimental runs which contradict the hypothesis the scientist is trying to demonstrate, or excessive filtration of “noise” which suggests a correlation where none can be shown to exist.’
All of the above clearly occurred, with the possible exception of “outright manufacture”. We’ll see how intellectually honest the CRU review is by EAU.