Development and criterion validity of a computerized text analysis measure of reflective functioning.

Area of Science:

• Psychological Assessment
• Computational Linguistics
• Personality Disorders

Background:

• Reflective Functioning (RF) assesses the capacity to understand subjective mental states.
• Traditional RF assessment is narrative-based and can be time-consuming.
• Developing automated methods for RF assessment is crucial for efficiency.

Purpose of the Study:

• To develop and validate a computerized text analysis version (CRF) of the Reflective Functioning scale.
• To explore the feasibility and criterion validity of the CRF measure.
• To identify linguistic markers associated with high Reflective Functioning.

Main Methods:

• Utilized 113 Adult Attachment Interviews (AAI) from clinical and non-clinical samples.
• Employed the Marker Approach to identify 54 linguistic markers of high RF language.
• Correlated CRF scores with established RF scores in Borderline Personality Disorder (BPD) patients and a non-clinical adult sample.

Main Results:

• Significant associations were found between the CRF and RF scales in both clinical (Spearman rho=.57, p<.0001) and non-clinical (Spearman rho=.57, p=.002) samples.
• The identified linguistic markers demonstrated strong correlations with Reflective Functioning.
• Preliminary criterion validity of the CRF measure was established.

Conclusions:

• A computerized Reflective Functioning (CRF) rating scale is feasible.
• The CRF measure demonstrates preliminary criterion validity.
• The CRF has potential to significantly enhance the efficiency of RF assessment.