Reliability, validity and standardization are well‐established principles in most areas of psychometrics, but are rarely mentioned in performance assessment within human psychopharmacology. Test reliability and standardization are examined here, while validity is covered in two succeeding articles. The undocumented reliability of human psychopharmacology performance tests makes the interpretation of findings difficult and ambiguous. Reliability, or consistency within and between test sessions is, however, easy to calculate. Several procedures for calculating reliability are described, with test–retest reliability recommended as the most appropriate single summary measure. Poor test standardization is another major problem, with many research groups using different tests. The extent of this problem is examined, and a set of standardization requirements proposed. They could comprise the basis for test manuals. The problems of undocumented reliability and unstandardized tests have been recognized within the field for many years, yet no moves have been made to remedy the situation. One simple solution would be for psychopharmacology journals to accept only documented tests, and to request test–retest reliability data.
No takes yet. Share an insight, caveat, or question.
A. C. Parrott (1991) studied this question.