Psychometric Properties of the Turkish Working Alliance Inventory-Short Form

This study aimed to build and examine the Turkish Working Alliance Inventory (WAI) short form which was originally developed by Horvath and Greenberg (1989) based on Bordin's pan-theoretical conceptualization of therapeutic alliance.

The Hacettepe University Psychotherapy Research Laboratory data set which included the masters's or doctoral level clinical psychology practitioners- patient pairs (83 sessions from 58 therapist and 83 patient). Short forms (e.g. therapist-client versions) that were built based on Tracey and Kokotovic's (1989) and Hatcher and Gillaspy's (2006) studies were analyzed with confirmatory factor analysis approach. During these analyses, the WAI measurements from the third sessions were based.

Confirmatory factor analysis results showed that the short forms which were suggested in the aforementioned literature, had good fit for both therapist and patient sample. Correlations between original Turkish forms and subscales, internal consistency levels and fit indices were reviewed together to reach a final decision. The short form which was suggested by Tracey and Kokotovic showed more appropriate statistical results. The final forms' Cronbach Alpha's levels were 0.90 and 0.86 for therapist and patient forms respectively.

Considering consistency with the observer form items (Soygüt and Uluç 2009) addition to inter-correlations, internal consistency levels and fit indices, the WAI-Short Forms were built    for therapist and patient. Our findings suggested that the WAI-Short Forms for therapist and patient have acceptable levels of reliability and validity for clinical and research settings in Turkey.