Malaysian Journal of Public Health Medicine 2016, Vol. 16 (3): 219-226 ORIGINAL ARTICLE THE GOODNESS-OF-FIT OF DASS-21MODELS AMONG UNIVERSITY STUDENTS Talwar P 1 , Tan Kock Wah 1 , Kartini Abd Ghani 1 , Nur Fatihah Mat Yusoff 1 1 Faculty of Cognitive Sciences and Human Development, University Malaysia Sarawak, 94300 Kota Samarahan, Sarawak ABSTRACT Depression, anxiety and stress are the most prevalent mental health problems among university students, which may adversely influence their psychological wellbeing. Considering these problems, the Depression, Anxiety and Stress Scale-21 (DASS-21) may hold more promise as a screening tool. The present study sought to assess competing models of the latent factor structure of the DASS-21 utilizing Confirmatory Factor Analysis, and also to examine the multi group invariance. An exploratory cross-sectional study was conducted among undergraduate students recruited from a public university in Sarawak, Malaysia. The self-report English language questionnaire was administered to 254 students. SPSS version 21 was used for data analysis. Overall, the DASS-21 exhibited adequate reliability, with the three-factor model showing good fit to the data from the current study. Gender invariance was also adequately demonstrated. However, high inter correlations indicated that depression, anxiety and stress may not be distinguishable constructs, as was originally perceived. Key Words: Confirmatory Factor Analysis, DASS-21, Gender invariance, University students INTRODUCTION Over the past few decades, there has been growing concern about the alarming rise in mental health issues, particularly among college and university students worldwide. During their college/university educational experience, students encounter both academic and non- academic stressors, which seem to inadvertently contribute to the decline in their mental health and may detrimentally affect their quality of life. Some of the common problems faced by students are issues related to their families, economic hardships, problems associated with adapting into the college environment, addictions, assignments and examinations. Although it cannot be construed that every student is at risk of developing mental health disorders, certain subgroups within the college population have a significantly higher prevalence of mental health problems, which is consistent with studies of the general population 1 . Undergraduate students are more vulnerable to psychological distress because their transition from high school to university student status, comes at a developmentally challenging phase when they are transitioning from adolescence to adulthood 2-5 . Hence, any decline in their mental wellbeing can have a profound impact on all aspects of campus life including at the individual level, the interpersonal level and even the institutional level 6 . Depression, anxiety and stress are the most prevalent mental health problems among the student population 7 . According to a recent survey conducted by the National College Health Assessment sponsored by the American College Health Association (ACHA-NCHA), which examined data from nearly 125,000 students from more than 150 colleges and universities, it was found that about one third of college students had difficulty functioning at some point in the last twelve months due to depression, and almost half reported they felt overwhelming anxiety in the past year 8 . Prevalence of these conditions have also been reported by several studies conducted among the student population in Malaysia 9-11 . Depression, anxiety and stress share an interdependent relationship and often influence one another 12-14 . Research from numerous investigations indicates that there is substantial overlap in anxiety and depressive symptoms and comorbid diagnoses in youth, yet these constructs can be adequately differentiated 15 . Explanation for the symptom overlap and diagnostic comorbidity has been greatly clarified by the tripartite model 16-18 . This model posits that anxiety and depression share a common component of negative effect, but can be differentiated by low positive affect associated with depression and high physiological hyperarousal associated with anxiety 16 . Although researchers have typically regarded anxiety and depression to be conceptually distinct, the distinguishability of these constructs has been difficult during screening 16, 17 . Considering these problems, the Depression, Anxiety and Stress Scales (DASS) 12 may hold more promise as a screening tool, since it was specifically designed to effectively maximize discrimination between the distinctive features of depression, anxiety and stress, and to minimize measurement of what these states have in common. The DASS comprises of three subscales. The DASS Depression Scale assesses anhedonia, devaluation of life, dysphoria, hopelessness, inertia, lack of interest or