Internal validity can be threatened by:
Bias (bias in collection, statistical analysis or interpretation)
Random Error
External Validity can be threatened by:
External validity is dependent on internal validity. Therefore, it is important to address any issues of internal validity in the study design.
Insufficient external validity may make the study results less relevant for application in clinical practice or in the general population.
Applicability issues: even if effectiveness is established, this may not always mean it is widely applicable.
(Spieth et al., 2016)