Modeling the population dynamics: Life expectancy of Malaysians

Kwa, Bee Yan (2007) Modeling the population dynamics: Life expectancy of Malaysians. Universiti Malaysia Sabah. (Unpublished)


Download (2MB) | Preview


The purpose of this study is to examine the factors that affect the life expectancy of Malaysians and attain the best model in order to represent the relationship. The trend of Malaysians' life expectancy is increasing from year to year and women generally have higher life expectancy compared to men. The six factors included in this project are crude death rate (CDR), labor force, gross domestic product (GDP), total fertility rate (TFR), population and the number of road accidents. All the data were taken from year 1980 to year 2005 from the Department of Statistics and the statistical analysis was done by using SPSS version 12.0 and Microsoft Excel. The all-possible-subset method with interactions is used for selecting the best set of predictor variables for multiple regression. The correlation coefficient between the variables and life expectancy were strongly correlated (p≈0.00) with linear relationship. There were a total of 120 possible models from the combination of all the variables and first order interaction. The models were selected when all the variables' significant values were less than 0.05 (p-value < 0.05). The eight criteria test is then carried out for 75 selected models to obtain the best model. The outcome of the study proved that total fertility rate (TFR) has a significant role towards the life expectancy of Malaysians. The best model contains eight parameters with the sum of square error (SSE) is equal to 0.21223575. The residual test showed that the relationship between the observed and prediction value is linear. The randomness test for the residual indicated that the residual is normally distributed and independent. The best model that is obtained from this study has perfectly predicted our observed value. With the evidence of the residual and randomness test, it is therefore can be concluded that the model is a good model within the range of this study.

Item Type: Academic Exercise
Uncontrolled Keywords: life expectancy,gender, labor force, gross domestic product, total fertility rate, crude death rate, population, number of road accident
Subjects: H Social Sciences > HM Sociology
Divisions: SCHOOL > School of Science and Technology
Date Deposited: 11 Jul 2013 05:57
Last Modified: 23 Oct 2017 07:05

Actions (login required)

View Item View Item