What are the four types of validity in Shadish and colleagues' taxonomy, and what does each type concern?

Shadish and colleagues' taxonomy identifies four types of validity: statistical conclusion validity, internal validity, construct validity, and external validity. Statistical conclusion validity concerns whether an empirical relationship between a presumed cause and effect truly exists. Internal validity concerns whether the independent variable, rather than some other factor, caused the outcome. Construct validity concerns whether the observed persons, settings, and cause-and-effect operations adequately represent the underlying constructs. External validity concerns whether causal inferences generalize across variations in persons, settings, time, and outcome measures.

The four types of validity in Shadish and colleagues' (2002) taxonomy are statistical conclusion validity, internal validity, construct validity, and external validity. Each addresses a distinct aspect of inference in quantitative research. Statistical conclusion validity is about the inference that there truly is an empirical relationship, or correlation, between the presumed cause and the effect; researchers must provide strong evidence that an observed relationship is real. Internal validity concerns the inference that, if an empirical relationship exists, it is the independent variable rather than something else that caused the outcome, so researchers must rule out competing explanations. Construct validity concerns inferences from the observed persons, settings, and cause-and-effect operations included in a study to the higher-order constructs these instances might represent; this includes whether an intervention adequately represents the underlying causal construct and whether outcome measures are good operationalizations of their intended constructs. External validity concerns whether inferences about observed relationships hold over variations in persons, setting, time, or measures of the outcomes, meaning it addresses the generalizability of causal inferences, which is critical for evidence-based nursing practice. The source text notes that each type of validity is threatened by different factors and that design decisions often involve trade-offs among them, particularly between internal and external validity.