Forecasting Deterioration of Mental Health during COVID-19 Pandemic and Lockdown - Examples from NepalJyoti U. Devkota 1, 1Department of Mathematics, Kathmandu University, Dhulikhel, Kavre, Nepal, e-mail : [email protected] ID: 0000-0001-5293-1753 

Here, mental health during COVID-19 pandemic and lockdown is studied. Multivariate statistics techniques namely, Canonical Correlation Analysis (CCA) and Structural Equation Modeling (SEM) is used. This study is based on a primary data. This data has information on 38 variables related to mental health. Some of the variables used in SEM are, Worry about food, Eat limited variety of food, Eat food that you did not want to eat, Eat smaller meal, Eat fewer meal, No food to eat of any kind, Affected my routine, Limited my social life, Reduced my travelling, Increased my household food consumption, Increased food shortage, Disturbed my sleeping pattern, Increased financial uncertainty, Decreased earning and income, Had fears of losing my job, Depressed due to uncertainty, Anxious due to uncertainty, Afraid of getting COVID-19, Afraid of losing my life and my relatives life, New and Media increased my tension, I am afraid of COVID outbreak and Stress level and anxiety level increased. Here three models were tested. The best model gave top values in terms of model efficiency parameters. This model is also represented with the help of path diagram. CCA predicted the relationship between the two groups of variables. These two groups are Cause and Effect of Deterioration in Mental Health. In SEM and CCA, model validation was done by drawing random samples of size 400.