Validity

 

           Validity refers to the accuracy of a study’s measurement tools or results. That is, it describes how likely a statement or measurement reflects the truth. Internal validity refers to the accuracy of an association or causal link and external validity refers to the generalizability of the measurement or data. For example, if a study finds that 90% of its pregnant respondents drank alcohol and concluded that pregnancy drinking rates must be that high, we may assume that the finding is not very externally valid as the data is not generalizable to the larger population (there may have been a flaw in the study design that resulted in this finding). If the 90% was found because interviewers asked participants if they had ever had alcohol rather than if they had alcohol during pregnancy, the measurement would not be internally valid because it does not accurately describe the situation. Chapter 16 of the Aschengrau and Seage text discusses validity in screening tests.