The different versions of the state-trait anxiety inventory

Dear Editor,

This letter attempts to bring attention to a possible source of error on the use and citation of one of the most often used questionnaires in our professional practice, the State-Trait Anxiety Inventory (STAI).

The anxiety inventory developed by Spielberger comes in 2 versions: the “STAI-form X”¹ version and the later “STAI-form Y” version.² In this second form, 37% of the original items are changed for other, simpler ones, improving the factorial structure of the scale. For example, in the “STAI-form X” original version, Item 23 indicates “I feel like crying”, while in the “STAI-form Y” version the item reads “I feel satisfied with myself”. As we see in this example, in some cases it is not only the content of the item that changes, but also the directionality of the score. Consequently, the correction is the scale is different and the results obtained using the 2 versions are not comparable.

Insofar as using the STAI in research studies, it can be seen that the version utilised in international contexts is the “STAI-form Y”. However, in Spain the one used the most is the Sin embargo, “STAI-form X”.³

This is possibly due to the greater diffusion that the “STAI-form X” has in our country, continuing up-to-date with its eighth version⁴ and available commercially through TEA Ediciones, S.A. The psychometric properties of the test were even revealed recently.⁵ The inventory enjoys great acceptance among professionals because of its simple administration and good psychometric properties. It should be pointed out that a Spanish translation of the Y form exists, available at Mapi Research Trust (http://www.mapi-trust.org), but it does not have published psychometric data.

Problems can arise in international multicentre studies, or when comparisons between results are performed without considering the fact that different versions might be involved, or if the exact version used is not specified.

This situation occurred in a multicentre study carried out at the Institute of Psychiatric Research (Instituto de Investigaciones Psiquiátricas, Bilbao-Fundación M. Josefa Recio). The focus was on attempting to evaluate the efficacy of the psycho-educational intervention, “EDUCA-III” (IRSCTN: 32545295. IP: M. Martín Carrasco), on the overload of caregivers of individuals with severe mental illness, with the participation of Spanish and Portuguese populations. One of the result variables was measured using the STAI. However, while the case report form (CRF) in the Spanish sample used the X form, in the Portuguese CRF the Y form was inadvertently used. When this fact was discovered, in spite of having gathered data from both populations, the sample data from the smaller group, of Portuguese origin, had to be eliminated from the analysis of the variable.

For all these reasons, we recommend that you should always check the version of the scale being used and should identify it appropriately in any publication. Likewise, it would be a good idea to have the assessment data and psychometric properties of the “STAI-form Y” available in Spanish.