a data point that is more than two standards deviations from the mean. Removal of outliers without articulating a legitimate reason may constitute data, falsification.

Practices and policies related to recording, storing, auditing, archiving, analyzing, interpreting, sharing, and publishing data.

use of statistical methods to fill in or replace missing or lost data. Imputation is not considered to be fabrication if it is done honestly and appropriately.

a committee that monitors data from human subjects research to protect participants from harm and promote their welfare. DSMBs may recommend to an institutional review board that a study be stopped or altered.

See Audit.

recorded information used to test scientific hypotheses or theories. Data may include laboratory notebooks (paper or digital), field notes, transcribed interviews, spreadsheets, digital images, x-ray photographs, audio or video recordings, and outputs from machines (such as gas chromatographs or DNA sequencers). Original (or primary data) is drawn directly from the data source; secondary (or derived) data is based on the primary data.

the idea that the institutional culture plays a key role in preventing research misconduct and promoting research integrity. Strategies to promote a culture of integrity include education and mentoring in the responsible conduct of research; research policy development; institutional support for research ethics oversight, consultation, and curriculum development; and ethical leadership.

fixing a minor problem with a published paper. A minor problem is one that does not impact the reliability or integrity of the data or results. Journals publish correction notices and identify corrected papers in electronic databases to alert the scientific community to problems with the paper. See also Retraction.

a right, granted by a government, which prohibits unauthorized copying, performance, or alteration of creative works. Copyright laws include a fair use exemption which allows limited, unauthorized uses for non-commercial purposes.

in human subjects research, subsequent review of a study after it has been approved by an IRB. Continuing review usually happens on an annual basis.