Recently a paper was published that had the aim of validating a borderline personality disorder instrument1 which gave excellent information on the measurement of this personality disorder. Nevertheless, a brief remark should be made about the concept of construct validity used by the authors of the paper, which leads to an interesting debate.
Validation is the process by which a measurement instrument is shown to be valid. Validity is classically therefore held to be “that which measures what it aims to measure”.2 However, things are not as simple as they sound. The concept of validity has changed significantly during history3: these changes started in the 1950s in disciplines close to the area of healthcare, and they were drawn up by technical committees who were designing regulatory manuals.4 The area of healthcare is now included in this initiative, and there is a guide to selecting measuring instruments within this field.5
Validity is currently understood to consist of the degree to which evidence and theory make it possible to interpret the scores of a test that was created with a purpose,6 and based on this definition five sources of evidence are postulated: validity based on the relationship with other variables, on response processes, on the consequences, and on internal content and structure.6 The latter source of validity refers to the use of factorial analysis as the means of checking the amount of factors and whether an item pertains to a factor.
It is due to this reason that an instrument is not validated in itself, but rather that validation takes place of the interpretations arising from the instrument. Note that the paper in question talks of validation of the construct, when the correct thing would be to talk about validity based on the construct (in contemporary classification validity based on internal structure). Finally, this difference seems to be trivial and insignificant; nevertheless, it contains a modification in how the concept of validity is understood, a viewpoint that has recently arisen as a debate in the area of healthcare.7
FinancingNone.
Conflict of interestsNone.
Please cite this article as: Ventura-León JL. ¿Validez de constructo o validez basada en el constructo?: comentarios a Soler et al. Rev Psiquiatr Salud Ment (Barc). 2017;10:221.