Pseudoscience is a proposition, a finding or a system of explanation that is presented as science but that lacks the rigor essential to the scientific method. Pseudoscience can also be the result of research that is based on faulty premises, a flawed experimental design or bad data.
The term pseudoscience can refer to a single claim or statement that is purported to be backed by science or data but doesn't stand up under scientific scrutiny. A pseudoscience may also be a complex system, such as astrology, which purports to explain events in the world as brought about and affected by astronomic phenomena. Like astrology, many pseudosciences are relatively harmless. Others, however, may be used to provide "scientific" support for unethical behavior. Physiognomy, for example, is a false science dating back to at least 500 BC that mistakenly correlates physical characteristics with personal traits. Even in modern times, adherents of physiognomy apply its theories to justify and promote inequality and racial profiling.
Pseudoscience may be offered in good faith, simply as a result of misinformation or poor analysis of data. However, scientific methods can also be intentionally applied in a flawed manner to create unwarranted confidence in conclusions that would not be supported by a more rigorous approach. In this case, pseudoscience is sometimes used to promulgate disinformation.
Within IT, data science is particularly vulnerable to flawed assumptions, methods and interpretation. It's sometimes assumed that big data analytics, for example, always yields valid information simply by virtue of the volume of data accessed, and that algorithms are similarly reliable. In both cases, however, the validity of the outcome can only be ensured by the quality of the input and the methods applied to it.
See also: data-driven disaster, falsifiability