Following content validation, more formal statistical procedures are used to assess the validity of a measurement. Here, a distinction may be drawn between measures of concepts for which there exists some type of criterion with which the measure can be compared, and those for which no criterion exists. The former include screening and diagnostic tests and predictive measures. The latter group include measures of abstract concepts such as quality of life, happiness, or disability. Validation procedures for the first group are relatively straightforward and include variants of criterion validation. Validation procedures for measures of the second type involve construct validation that is more extensive than criterion validation. These two approaches will next be described in sequence.