Extracting principal components and determining critical questions in conners ADHD questionnaire

Conners Adult ADHD Rating Scale (CAARS) is one of the reliable questionnaires in evaluating ADHD in adults. The goal of this study is to extract principal components of the screening version of questionnaire and evaluate the validity of missed answers estimation. This study is performed on 380 volunteers. Eight principal components are extracted by means of PCA. In the next step, for each question, the answer estimation is calculated (assuming this answer was missed), and then a Kruskal-Wallis test is performed to evaluate the difference between the original answer and its estimation. Results indicate that for some particular questions there is a significant difference between the original and estimated answers. However, result of Multiple Comparison Procedure shows that this estimation when evaluated in the whole group does not have any significant difference with the original value in any of the questionnaire four subscales. Notwithstanding this fact, if the missed question is a critical one, more considerations shall be paid.