Fabricating Data

Fabricating data is the process of producing imaginary data to support a certain hypothesis [cite, cite]. In this case, the Researcher comes up with datapoint and inject them into actual measurements in order to achieve or legitimize a certain conclusion. Data fabrication is neither defensible nor prevalent [cite, cite]; however unfortunately it can be deployed by researchers.

Here we focus on two main known/speculated approaches: data generation, data duplication.

• Data Generation is the process of generating completely new datapoint with the goal of steering test statistics or any other metrics toward a satisfactory conclusion.
• Data Duplication is the process of duplicating certain datapoint in a selected group, in order to weight the group in certain direction.

Like many other hacking strategies, a range of options can be used to customize the behavior of the Researcher during this hacking strategy.

---

Last update: 2021-09-18