The validity and reliability of Turkish version of "barkley adult attention deficit hyperactivity disorder rating scale-IV (BAARS-IV) current symptoms self report subscale" in students of medicine

Aim and Hypothesis: Our aim is to investigate the psychometric properties of the 'BAARS-IV Self Report, Current Symptoms'- Turkish version. Method: A total of 417 participants completed the Sociodemographic Information, Beck Depression Inventory, Beck Anxiety Inventory, ASRS and BAARS-IV-Self Report-Current Symptoms questionnaires. 402 participants were included to analysis. Exploratory and confirmatory factor analysis performed for structural validity. For the convergent and divergent validity for SCT scale, EFA and CFA were conducted with inattention and hyperactivity/impulsivity items. In addition, correlation analysis performed with total scores of each subscales. For reliability, test-retest, internal consistency, split-half reliability statistics are performed. We also conducted parallel forms reliability for the ADHD part of the scale. Test-retest reliability analysis was performed with some of the participants who completed the questionnaire 12-22 days after the first questionnaire, by using Intraclass-correlation Coefficient with absolute agreement and two-way mixed options. Results: SCT item-9 is excluded from the scale because of low communality and low corrected item-scale correlation. EFA and CFA results are acceptable. SCT item 6,7,8 showed better convergent and discriminant validity. Split half reliability, internal consistency results are acceptable. Test-retest statistics showed acceptable reliability for hyperactivity subscale and good reliability for the rest of subscales. Conclusion: This study suggests that the BAARS-IV Self Report Current Symptoms Scale is a reliable and valid measure of ADHD and SCT. We found that SCT items 6,7,8 are more specific. Item 15 in ADHD scale - 'talks a lot in social situations'- showed some problems in internal consistency. We recommend that, item-15 in ADHD scale should be explained more specific way, at least in Turkish population. More research is needed to understand psychometric properties of the scale, especially in community and clinical populations.