Establishing a common metric to facilitate the understanding of test scores in Routine Outcome Monitoring applied to measurement instruments for anxiety

Background: Questionnaires for anxiety disorders come in different lengths, scales and with varying interpretation guidelines. This may hamper therapist-patient communication when discussing the score and the interpretation of where a patient can be placed on a severity scale. Expressing scores on a standardized common metric can facilitate the communication between therapist and patient. This thesis aims for a method that enables an easier interpretation of scores and in addition produces scores with a normal distribution. Method: Using the data of four anxiety questionnaires, namely the Brief Scale for Anxiety (BSA), the PADUA Inventory Revised (PI-R), the Panic Appraisal Inventory (POL/PAI) and the Impact of Events Scale Revised (IES-R), theta-based T-scores were calculated with the Item-Response Theory and deployed as a basis for crosswalk tables to look up T-scores from raw scores. Based on these crosswalk tables, transformation formulas were established to calculate T-scores....Show moreBackground: Questionnaires for anxiety disorders come in different lengths, scales and with varying interpretation guidelines. This may hamper therapist-patient communication when discussing the score and the interpretation of where a patient can be placed on a severity scale. Expressing scores on a standardized common metric can facilitate the communication between therapist and patient. This thesis aims for a method that enables an easier interpretation of scores and in addition produces scores with a normal distribution. Method: Using the data of four anxiety questionnaires, namely the Brief Scale for Anxiety (BSA), the PADUA Inventory Revised (PI-R), the Panic Appraisal Inventory (POL/PAI) and the Impact of Events Scale Revised (IES-R), theta-based T-scores were calculated with the Item-Response Theory and deployed as a basis for crosswalk tables to look up T-scores from raw scores. Based on these crosswalk tables, transformation formulas were established to calculate T-scores. To validate these, calculated T-scores were compared to theta-based T-scores. Results: Most of the calculated T-scores had a normal distribution and the correlations between both methods to arrive at the T-scores were significant, the highest correlation was found for the BSA and the IES-R. Discussion: Due to a significant correlation and a large sample size a new method to arrive at a common metric was established by linking every raw score on a T-score metric. This provides a way to facilitate the interpretation and discussion of outcome scores. Further, with this new method disorder severity can be calculated and looked up due to tangible cut-off scores.Show less

Leiden University Student Repository

Documents

In Collections

Establishing a common metric to facilitate the understanding of test scores in Routine Outcome Monitoring applied to measurement instruments for anxiety