Psychometric Properties of the Arabic Vanderbilt Children's ADHD Diagnostic Rating Scale (VADRS-A) in a Saudi Population Sample.

Abstract

Objective: This study aimed to utilize Principal Component Analysis (PCA) and Confirmatory Factor Analysis (CFA) to investigate the constructive validity of the Arabic translation of the Vanderbilt Attention Deficit/Hyperactive Disorder (ADHD) Diagnostic Scale (VADRS-A) using its two versions, the Arabic Vanderbilt ADHD Diagnostic Parent Rating Scale (VADPRS-A) and the Arabic Vanderbilt ADHD Diagnostic Teacher Rating Scale (VADTRS-A).

Method: A descriptive research design was employed. Children were assessed by their parents and teachers, and a cluster sample of 1812 participants was chosen from many schools in Saudi Arabia, divided into two groups: parents (504) and teachers (1308) of children whose ages ranged between 5 to 12 years old. The VADPRS-A and VADTRS-A were administered to parents and teachers under the supervision of the Saudi ADHD Society.

Results: The results of the PCA of VADPRS-A found that the six factors saturate one general factor that explained (59%) of the total variance of the factor matrix with eigenvalues (3.540). Similarly, the PCA of VADTRS-A demonstrated that the five factors were saturated on a general factor that explained (69.20%) of the total variance of the factor matrix with eigenvalues (3.460). Also, the results indicate the high internal consistency of VADPRS-A and VADTRS-A, all factors correlated together and the total scores positively and significantly statistically (p>.001) correlation coefficients ranged between (0.296 to 0.843) for VADPRS-A, and ranges between (0.432 to 0.939) for VADTRS-A. Also, the Cronbach's α coefficient values for the six factors and total score of VADPRS-A were (.906, .925, .900, .896, .853, .872, .959) respectively, and these values are close to the values of the McDonald's ω for the factors and the total score were (.908, .923, .901, .871, .850, .877, .925) respectively. In the same way, Cronbach's α coefficients were (.967, .921, .914, .858, .948, .971) for all factors and the total score of VADTRS-A respectively, and these values are close to the values of the McDonald's ω (.968, .921, .919, .856, .943, .965) for all factors and the total score of VADTRS-A. In addition CFA for VADPRS-A and VADTRS-A models showed acceptable factor loading and good values of goodness-of-fit indices; CFI, TLI, RMSEA, IFI, and GFI (0.956, 0.942, 0.049, 0.956, 0.952) respectively for VADPRS-A model, and were (0.958, 0.932, 0.051, 0.963, 0.964) for VADTRS-A model, all of these were at an acceptable range. These results suggest a fit with the previous theoretical literature about VADPRS and VADTRS and DSM-5 and ICD-11 criteria of ADHD.

Conclusion: These findings highlighted the good psychometric properties of VADRS-A in both its versions VADPRS-A and VADTRS-A in the Saudi environment. Due to these findings, we suggest utilizing VADPRS-A and VADTRS-A during ADHD diagnosis in children 5-12 years old in Saudi Arabia, to facilitate early diagnosis and intervention, and to help mitigate the risks of ADHD during subsequent developmental phases in children's lives.