If a finding is repeated, it is described as being reliable. Within the general meaning of the term, it is also used more specifically within psychological assessment and research.
For a research finding to be reliable, it must be shown to exist on successive investigations under the same condition (replication). For a psychometric assessment to be reliable, it should have both internal and external reliability. Answers to a questionnaire or inventory may be checked to see if respondents answer all questions in the same way or if they contradict themselves. This is a measure of internal reliability. Responses may also be checked over a period of time to see if there is stability of measurement over times. If respondent gives the same responses or obtains the same scores consistently over time, then the measure is said to have external reliability.