Alternative terms: publication bias, outcome reporting bias
Selective reporting bias may include::
(1) a failure to report either a failure or success (poor or good performance) for a therapy or diagnostic test
(2) "cleaning up the data" so that results look more convincing than they actually are
Reasons why selective reporting bias may occur:
(1) a granting agency or journal may prefer a winner rather than a loser
(2) a company may think that any negative information may harm the sale of a product
(3) a researcher may not want to report something that goes against the existing literature or which might be construed as a failure
(4) a researcher may want to climb on a popular bandwagon
(5) a clique in power only wants things published that support the party line
Consequences may include:
(1) Patients undergo a test or receive a therapy which does not perform as expected.
(2) Patients may undergo a therapy with serious side effects.
(3) Patients may not receive a therapy which would have been beneficial to them.
This form of bias needs to be distinguished from "reporting bias", which may used as a an alternative term for recall bias.
