The degree of external validity of a particular study or experiment refers to the extent to which the results can be generalised beyond the particular sample or circumstances of the original study or experiment. External validity is of particular concern to experiments involving human or animal subjects, since even if human subjects exhibit certain behaviours or reactions in an experimental setting, one must always ask whether they will behave in the same way in the real world, outside the artificially-imposed conditions of the experiment. Likewise, animal observation or experimentation generally is conducted on animals place in highly artificial environments, which may lead to quite different behaviours than are typically observed in their natural habitat. One way to improve the external validity of results is by combining experimental manipulations with real-world observational studies, and seeing whether these two different approaches yield similar results.
External validity: excellent introduction with a focus on psychological applications
External validity: useful introduction from Research Methods Knowledge Base
External validity definition and examples: brief introductory discussion including examples and methods for improving external validity
How to assess the external validity of therapeutic trials: a more advanced treatment from an epidemiological journal