Leveraging Artificial Intelligence for Stress, Anxiety, And Depression Assessment: An Empirical Study Using Dass-42 Among Postgraduate Management Students

Journal of Research and Development Peer Reviewed International, Open Access Journal. ISSN : 2230-9578 | Website: https://jrdrvb.org Volume-17, Issue-10(IV)| October2025 46 Leveraging Artificial Intelligence for Stress, Anxiety, And Depression Assessment: An Empirical Study Using Dass-42 Among Postgraduate Management Students Dr. Anup M. Mule1, Sagar Kurane2, Dr. N. D. Mangore3, Poonam Doke4, Vaishnavi Patil5 1Associate Professor, Department of Management, FoCM, SGU, Atigre. 2Assistant Professor, Department of Management, FoCM, SGU, Atigre. 3Assistant Professor, Department of Psychology, SCA&S College, Malwadi-Kotoli. 4-5PG management students, Department of Management, FoCM, SGU, Atigre. Manuscript ID: JRD -2025-171012 ISSN: 2230-9578 Volume 17 Issue 10(IV) Pp. 46-51 October 2025 Submitted: 22 Sept. 2025 Revised:05 Oct. 2025 Accepted:13 Oct. 2025 Published: 31 Oct. 2025 Abstract This research examines the psychological well-being of postgraduate management students at Sanjay Ghodawat University, Atigre, by integrating the Depression Anxiety Stress Scales (DASS-42) framework with the conceptual use of Artificial Intelligence (AI). The study involved 89 MBA students from both first and second year, allowing for an in-depth evaluation of mental health patterns in an academic context. Its key aims were to assess the levels of depression, anxiety, and stress among students, analyze differences across demographic factors such as gender and age, and conceptually investigate how AI could enhance psychological evaluation through automated classification, predictive analysis, and personalized intervention systems. By merging psychological insights with AI-based approaches, this study adds value to existing literature on mental health assessment and supports the growing field of digital mental health practices in higher education. Keywords: DASS-42, Artificial Intelligence (AI), Student Mental Health, Stress, Anxiety, Depression. Introduction Mental health concerns among higher-education students have been increasing in recent years. Graduate and professional students often face high workloads, performance pressures, and career uncertainty, which can elevate rates of anxiety, depression, and stress. Surveys have found that a substantial proportion of college students report moderate or severe symptoms of anxiety and depression during their studies. Business school students in particular encounter the challenges of rigorous academic schedules alongside volatile business environments and financial pressures. These aspects emphasize the need to evaluate mental health among MBA students. The Depression Anxiety Stress Scales (DASS-42) is a reliable and widely recognized self-assessment tool designed to identify negative emotional states. It includes three components—Depression, Anxiety, and Stress—each containing 14 questions rated on a 0 to 3 scale, resulting in subscale scores ranging from 0 to 42, where higher scores reflect greater emotional distress. In this study, the DASS-42 is applied to assess the psychological well-being of MBA students at SGU, addressing a significant research gap at the local level. 2. Rationale: MBA programs often attract students with high ambition and extracurricular commitments, which can compound stress. By measuring DASS-42 scores and examining demographic patterns, this research provides baseline data on the psychological well-being of these students. Identifying any vulnerable subgroups (by gender or age) could guide targeted interventions. Moreover, this study seeks to explore, at a conceptual level, how emerging AI analytics might enhance traditional assessment, providing more personalized insights into student mental health. 3. Scope: This research adopts a cross-sectional survey design conducted in 2025 among current MBA students at SGU. It centers on self-reported levels of depression, anxiety, and stress using the DASS-42 scale. Quick Response Code: Website: https://jrdrvb.org/ DOI: 10.5281/zenodo.17464074 Creative Commons (CC BY-NC-SA 4.0) This is an open access journal, and articles are distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License, which allows others to remix, tweak, and build upon the work noncommercially, as long as appropriate credit is given and the new creations ae licensed under the idential terms. Address for correspondence: Dr. Anup M. Mule, Associate Professor, Department of Management, FoCM, SGU, Atigre. ,How to cite this article: Dr. Anup M. Mule, Sagar Kurane, Dr. N. D. Mangore, Poonam Doke, Vaishnavi Patil, (2025). Leveraging Artificial Intelligence for Stress, Anxiety, And Depression Assessment: An Empirical Study Using Dass-42 Among Postgraduate Management Students Journal of Research & Development, 17(10(IV)), 46-51 Original Article Journal of Research and Development Peer Reviewed International, Open Access Journal. ISSN : 2230-9578 | Website: https://jrdrvb.org Volume-17, Issue-10(IV)| October2025 47 The study also considers demographic factors such as gender and age group to identify variations in psychological well-being. The scope is limited to one university’s MBA program and does not track changes over time or include clinical diagnoses. Interventions are not implemented; rather, data is analyzed to understand prevalence and group differences. The study also conceptually considers the application of AI techniques to this dataset, but does not implement machine learning algorithms. 4. Conceptual Framework We conceptualize psychological distress as comprising three related constructs: depression (feelings of low mood, worthlessness etc…), anxiety (excessive worry, arousal, and panic), and stress (perceived inability to cope and relax). The DASS-42 framework treats these as dimensional continuums of negative affect. In this framework, each MBA student’s DASS-42 scores reflect their current emotional state. Socio-demographic factors (e.g., gender, age, year of study) are potential correlates of these scores. Prior research suggests mixed results: some studies find higher anxiety in female students or higher stress in older students, while others find no gender differences. We will interpret our findings in light of this literature. In addition, we conceptually integrate the role of artificial intelligence (AI) in mental health analysis. AI methods like decision trees or logistic regression can classify students into risk categories (e.g., severe vs. mild symptoms) based on DASS scores and other data. Clustering algorithms could identify subgroups of students with similar symptom profiles or stressors. While this study does not perform such AI analyses, we discuss how these tools could complement classical statistical methods, potentially improving early detection and tailored support for at-risk students. 1. Research Objectives 2. To measure depression, anxiety, and stress levels among SGU MBA students using the DASS-42 instrument. 3. To compare DASS-42 scores by demographic subgroups. 4. To conceptually explore how AI-based methods could enhance analysis of the DASS-42 data and support personalized mental health interventions for students. 6. Hypothesis H0: There are no significant differences in depression, anxiety, and stress levels across demographic variables (gender, age). H1: There are significant differences in depression, anxiety, or stress levels across demographic variables (gender, age). 7. Research Approach A quantitative survey methodology was employed. The study used a non-experimental, descriptive-correlational approach. Primary data came from the DASS-42 survey. Data analysis involved descriptive statistics (means, standard deviations, frequencies) and inferential statistics to test for group differences. Specifically, independent-samples t-tests were planned for comparing male vs. female mean scores, and a one-way ANOVA was planned for comparing mean scores across three age categories. All tests used a significance level of α=0.05. Additionally, p-values were noted to determine statistical significance (with p<0.05 considered significant). As the AI objective is conceptual, no predictive modeling was conducted; however, statistical results were interpreted with an eye toward how AI classification or clustering could be later applied. 8. Population and Sample The population of interest is all MBA students at Sanjay Ghodawat University. The sample consisted of N = 89 MBA students who completed the DASS-42 survey. Among these, 48 identified as female and 41 as male. Age distribution was categorized into three groups: 18–21 years (n=19), 22–24 years (n=65), and 25–30 years (n=5). The small number in the oldest group is noted as a limitation, but was retained for analysis. Participants were selected via voluntary response; thus, the sample is effectively a convenience sample of those who chose to complete the survey. 9. Data Analysis The following section presents the demographic characteristics of the respondents who participated in the study. The analysis includes variables such as gender, location, family type, exposure to employment or family business, and age. Journal of Research and Development Peer Reviewed International, Open Access Journal. ISSN : 2230-9578 | Website: https://jrdrvb.org Volume-17, Issue-10(IV)| October2025 48 Understanding the demographic composition of the participants is crucial for contextualizing the findings related to depression, anxiety, and stress levels measured by the DASS-42 scale. Variable Category Frequency Percentage Gender Female 48 53.9% Gender Male 41 46.1% Location Rural 16 18.0% Location Semi-Urban 33 37.1% Location Urban 40 44.9% Family Type Joint 36 40.4% Family Type Nuclear 53 59.6% Exposure to Employment/Family Business No 29 32.6% Exposure to Employment/Family Business Yes 60 67.4% Age 18–21 19 21.3% Age 22–24 65 73.0% Age 25-30 05 05.7% The data presented above highlight the demographic distribution of the respondents. The sample is relatively balanced in terms of gender, with a slightly higher number of female participants. Most respondents belong to semi-urban and urban areas, indicating a predominance of participants from moderately or highly developed localities. The majority of respondents live in nuclear families, reflecting changing family structures in contemporary society. Additionally, a large proportion of participants reported prior exposure to employment or family business, which could influence their stress and coping mechanisms. Age-wise distribution shows that the majority are between 22 and 24 years, aligning with the early adulthood stage where academic and career-related pressures are typically high. DAS Analysis Particular AVG Score Interpretation Depression 9 Normal Anxiety 11 Mild Stress 13 Mild Depression Category Range Frequency % Normal 0--9 54 61 Mild 10--13 11 12 Moderate 14--20 15 17 Severe 21--27 7 8 Extremely Severe 28+ 2 2 89 100.0 Anxiety Category Range Frequency % Normal 0--7 37 42 Mild 8--9 4 4 Moderate 10--14 23 26 Severe 15--19 12 13 Extremely Severe 20+ 13 15 89 100.0 Stress Category Range Frequency % Normal 0--14 55 62 Mild 15--18 12 13 Moderate 19--25 13 15 Journal of Research and Development Peer Reviewed International, Open Access Journal. ISSN : 2230-9578 | Website: https://jrdrvb.org Volume-17, Issue-10(IV)| October2025 49 Severe 26-33 7 8 Extremely Severe 34+ 2 2 89 100.0 A majority (61%) of respondents show no significant symptoms of depression, suggesting good emotional stability in this dimension. However, 37% of participants report at least mild to moderate depressive symptoms, which indicates a small but notable proportion may require emotional or psychological support. Only 10% (severe and extremely severe) exhibit high depressive tendencies and may need professional attention. Anxiety levels are relatively higher compared to depression and stress. Less than half of the respondents (42%) fall in the normal category, while 58% experience some level of anxiety, ranging from mild to extremely severe. Notably, around 28% of participants (severe + extremely severe) show elevated anxiety levels, indicating a need for stress management or counseling interventions. Most respondents (62%) report normal stress levels, implying adequate coping mechanisms in daily life. However, about 38% experience mild to severe stress symptoms. Though not alarming, this indicates the presence of manageable stress levels that may be reduced through relaxation, time management, or mindfulness training. Descriptive statistics for DASS subscales are presented in Tables 1–2. Table 1 shows the mean (M) and standard deviation (SD) for Depression, Anxiety, and Stress by gender. Table 2 shows these statistics by age group. Table 1: DASS-42 scores by gender (N = 89) Gender n Depression (M ± SD) Anxiety (M ± SD) Stress (M ± SD) Male 41 7.90 ± 7.94 9.85 ± 8.94 11.55 ± 9.59 Female 48 8.98 ± 8.01 11.44 ± 8.23 14.27 ± 7.94 Formula used for independent t-test: t = (X ₁ - X ₂) / √ ((s₁²/n₁) + (s₂²/n₂)) Table 2: Mean, t-value and p-value Variable Mean (Male) Mean (Female) t-value p-value Depression 7.90 8.98 -0.632 0.529 (NS) Anxiety 9.85 11.44 -0.860 0.392 (NS) Stress 11.55 14.27 -1.431 0.156 (NS) Interpretation: As p > 0.05 for all subscales, there is no significant difference between male and female students. The overall mean scores (combining both genders) were Depression = 8.56 (SD ≈ 7.94), Anxiety = 10.88 (SD ≈ 8.63), Stress = 13.19 (SD ≈ 8.86). These averages fall in the mild-to-moderate range on DASS severity scales. Table 3: DASS-42 scores by age group (N = 89) Age Group (years) n Depression (M ± SD) Anxiety (M ± SD) Stress (M ± SD) 18–21 19 9.47 ± 7.64 11.95 ± 9.71 14.21 ± 9.71 22–24 65 7.94 ± 7.84 9.78 ± 7.78 12.38 ± 8.44 25–30 5 13.20 ± 10.06 21.00 ± 9.67 19.80 ± 9.58 Interpretation: Older students (25–30) had noticeably higher mean scores, especially on anxiety and stress, but their group size is small (n=5). Table 4: Gender comparisons (t-test) for DASS subscales Subscale t p Interpretation Depression -0.632 0.529 Not significant (p>0.05) Anxiety -0.860 0.392 Not significant (p>0.05) Stress -1.431 0.156 Not significant (p>0.05) All p-values exceed 0.05, indicating no statistically significant gender differences for depression, anxiety, or stress. (Negative t-values indicate females had slightly higher means than males for all subscales, but differences were nonsignificant.) Table 5: Age Differences (ANOVA): One-way ANOVA was conducted for each subscale across the three age groups. Formula used for One-way ANOVA: F = MSB / MSW Where MSB = SSB/df₁ and MSW = SSW/df₂ Journal of Research and Development Peer Reviewed International, Open Access Journal. ISSN : 2230-9578 | Website: https://jrdrvb.org Volume-17, Issue-10(IV)| October2025 50 Subscale F p Interpretation Depression 1.185 0.311 Not significant (p>0.05) Anxiety 4.424 0.0148 Significant (p<0.05) Stress 1.820 0.168 Not significant (p>0.05) Interpretation: Anxiety levels differ significantly among age groups (p = 0.0148). Post-hoc analysis revealed that the age group 25–30 years scored significantly higher on anxiety than 22–24 years. The ANOVA shows that only anxiety scores differed significantly by age group (F ≈ 4.42, p ≈ 0.0148). Post-hoc comparisons (Tukey test) indicated that the 25–30 age group had significantly higher anxiety than the 22–24 group (mean difference ≈ 11.22, p ≈ 0.013). No significant pairwise differences were found for depression or stress. In summary, most comparisons yielded p>0.05 (Table 3–4), so the overall null hypothesis (H0) of no differences across gender and age is not rejected for depression and stress. The exception is that anxiety varies by age (H0 rejected for anxiety by age). Because of these findings, H1 is supported only in the domain of anxiety across age categories. Gender had no significant impact on DASS subscale scores. All statistical results are reported with degrees of freedom and p-values where applicable. Figures illustrating these results (e.g., mean scores by group) are not embedded here, but the data trends are summarized above. Discussion The present study aimed to measure the levels of depression, anxiety, and stress among postgraduate management students using the DASS-42 scale and to find out whether these levels differ based on factors like gender, age, and year of study. Additionally, it conceptually explored how Artificial Intelligence (AI) could be applied to improve the precision and interpretation of mental health assessments. The findings provide valuable insights into the psychological state of management students and highlight potential avenues for incorporating AI-based analysis into educational and counseling frameworks. The analysis showed that most postgraduate management students experienced moderate levels of depression, anxiety, and stress, with some individual differences. When comparing these factors across gender, age, and year of study, no major differences were found in depression and stress levels. This suggests that such emotional challenges may result from common academic and personal pressures rather than demographic factors. However, anxiety levels did vary by age—students aged 25 to 30 reported higher anxiety than younger ones. This finding supports earlier studies indicating that older students often feel more anxious as they juggle studies, work, and family responsibilities. Thus, while stress and depression were similar across groups, anxiety seemed more influenced by age and life circumstances. These findings carry important implications for promoting mental well-being in management education. Colleges and universities should consider developing targeted programs to help reduce anxiety among older students, while also offering general wellness initiatives for the entire student body. The lack of significant gender differences challenges the common assumption that female students experience higher emotional distress, suggesting that both male and female management students face similar levels of academic and psychological pressure in today’s learning environment. Beyond the empirical outcomes, this research also conceptually examined the role of AI as a transformative tool in psychological assessment. Traditional use of tools like DASS-42 depends on fixed scoring criteria and manual interpretation, which may overlook complex emotional patterns. AI-driven techniques, such as machine learning classification and predictive modeling, hold promise for uncovering hidden distress profiles, tracking emotional changes over time, and identifying at-risk individuals early. For example, unsupervised algorithms could automatically group students based on underlying stress indicators that conventional statistics might miss. In this way, AI can act as a supportive layer to traditional psychometric evaluations, enhancing accuracy and contextual understanding. It is important to highlight that the AI aspect of this study was purely theoretical, and no practical development or testing of AI models was carried out. Therefore, conclusions related to AI applications should be viewed as future directions rather than experimental results. Subsequent studies could build upon this idea by training supervised models on extensive DASS datasets to classify psychological severity or by using natural language processing (NLP) to analyze students’ written reflections for emotional cues. Such approaches could facilitate more sensitive, real-time monitoring of mental health trends among students. This study, however, is not without limitations. The sample size was relatively small (n = 89) and limited to one university, which restricts the generalizability of the findings. Moreover, as a cross-sectional study, it cannot determine causation between demographic characteristics and psychological outcomes. Future research should adopt longitudinal designs to track how mental health indicators evolve over time and how AI-based systems could be practically implemented to predict and manage these patterns more effectively. Journal of Research and Development Peer Reviewed International, Open Access Journal. ISSN : 2230-9578 | Website: https://jrdrvb.org Volume-17, Issue-10(IV)| October2025 51 Conclusion All comparisons, except for anxiety across age groups, showed p-values above 0.05, suggesting no statistically significant differences. However, since age was found to have a significant impact on anxiety levels, the null hypothesis is rejected. This indicates that anxiety levels vary notably among postgraduate management students of different age groups, while no significant variations were observed in depression and stress with respect to other demographic factors. Accomplishments of study objectives: Objective Status Remarks 1. Achieved DASS-42 descriptive statistics. 2. Achieved Parametric tests. 3. Achieved Conceptual This study provides a comprehensive assessment of mental health indicators among MBA students at SGU. Key findings are: (1) overall moderate levels of depression, anxiety, and stress, (2) no significant gender differences in DASS-42 scores, and (3) a significant difference in anxiety by age, with older students exhibiting higher anxiety. These results suggest that MBA programs should attend to the mental well-being of all students, with particular focus on anxiety management for older cohorts. Integrating AI methods in future work could refine these insights and support personalized interventions, though current results already highlight where student support is most needed. Limitations and Future Work Limitations include the relatively small and self-selected sample (n=89), which may not generalize to all MBA students or other universities. The cross-sectional design captures only a single time point; mental health may fluctuate over the semester. The oldest age group had only 5 students, so anxiety findings by age should be interpreted cautiously. Future research could use larger samples, longitudinal tracking, and include additional variables (e.g., academic performance, employment status). Incorporating qualitative data (e.g., interviews) could enrich understanding of the sources of stress. Finally, implementing and empirically testing AI-based analytic tools on the DASS data would be a valuable next step, as suggested in the conceptual framework. References 1. Dehbozorgi, R., Zangeneh, S., Khooshab, E., Hafezi Nia, D., Hanif, H. R., Samian, P., Yousefi, M., &Lohrasebi, F. (2025). The application of artificial intelligence in the field of mental health: A systematic review. BMC Psychiatry, 25, Article 132. 2. Haruna, U., Mohammed, A.-R., & Braimah, M. (2025). Understanding the burden of depression, anxiety and stress among first-year undergraduate students. BMC Psychiatry, 25, 632. (Supports Introduction/Discussion: prevalence and gender differences in student bmcpsychiatry.biomedcentral.com.) 3. Karmakar, N., Saha, J., Datta, A., Nag, K., & Tripura, K. (2021). A comparative study on depression, anxiety, and stress among medical and engineering college students in North-East India. CHRISMED Journal of Health and Research, 8(3), 193–199. (Supports Introduction/Discussion: use of DASS-42 in Indian student samples and high distress journals.lww.com.) 4. Basha, S. E., Gull, M., Alquqa, E. K., Mahmoud, K., &Harhash, A. (2025). The role of AI in university students’ mental health: A bibliometric review. Discover Social Science and Health, 5, 131. (Supports AI Conceptual Framework: reviews AI/ML applications for student mental health, noting high accuracy of ML link.springer.com.) 5. Varghese, M. A., Sharma, P., & Patwardhan, M. (2024). Public perception on AI-driven mental health interventions: Survey research. JMIR Formative Research, 8(10), e64380. (Supports AI Context and Limitations: Indian attitudes toward AI in mental health, noting accessibility benefits and trust researchgate.net.)