| Abstract [eng] |
The object of this master’s thesis is statistical and demographic data about the number of people suffering from mental and behavioral disorders (hereinafter referred to as ‘‘PMBD“) in Lithuania over the period from 2006 to 2019. Data was taken from the periodical health statistics publications ‘‘Health of the Lithuanian Population and Activities of Health care institutions“, published by Health Information Centre of the Institute of Hygiene. The aim of the work is to investigate the dynamics of the number of patients with mental and behavioral disorders by statistical methods during the period under consideration in Lithuania. The tasks set to achieve the goal are: first of all, to investigate the dependence of the number of Lithuanian patients with mental and behavioral disorders on the population using correlational and regression analysis, then applying statistical hypotheses about the meanings of the parameters of the polynomial distribution, to determine whether the number of patients of the 7 disease-groups under consideration changes dramatically after a certain period of time, if so, indicated after some period (after a few years). Continuing the study by disorder to investigate the dynamics of changes in the number of patients annually and every four years, to describe and depict it visually and then using a series of criteria on the randomness and independence of the data, to determine which data of patients with mental and behavioral disorders can be used time-series methods. Lastly, using the time series equalization method, determine which data rows have a trend and which do not, using time series methods and to find estimates of parameters for the AR(1) and AR(2) models when using and without using the trend. After an analysis of the data of patients with mental and behavioral disorders and the population of Lithuania it was established that the PMBD figures correlate with the population figures of Lithuania. During the period from 2006 to 2019 the number of sick people strongly depends on the population of Lithuania (hypothesis H_0:ρ=-0.9, with a probability of 0.95 is adopted). This dependence is linear and can be written in the regression equation: y = 437539 - 0,0757 ∙ x + ε,ε~N(0; 1444406,189), here 𝑦 – number of PMBD, x – population. The forecast range for this equation for 2019 is y ∈[212108,82; 238259,94]. After applying the statistical hypothesis on the values of the polynomial distribution parameter it was found that the number of people in the PMBD of the 7 disease groups under consideration changes statistically significantly only every four years and this change does not yet appear after one or two years. This is also reflected in the color variation table of the dynamics of PMBD changes. The series criterion on data randomness and independence identified which data groups are subject to time series methods. When using the time series equalization method it was determined that all selected data rows have a trend, estimates of parameters for the AR(1) and AR(2) models were found when the trend is not eliminated from the data rows and when the trend is eliminated. |
