Homework answers / question archive / Xiang Cui Shangyi Guo Research Proposal Name University Name Course Feb 19, 2021 Preliminary Title (Factors Affecting the Life Expectancy of an Individual) Life expectancy can be defined as the number of years the individual can expect to live

Xiang Cui Shangyi Guo Research Proposal Name University Name Course Feb 19, 2021 Preliminary Title (Factors Affecting the Life Expectancy of an Individual) Life expectancy can be defined as the number of years the individual can expect to live

Sociology

Share With

Xiang Cui Shangyi Guo Research Proposal Name University Name Course Feb 19, 2021 Preliminary Title (Factors Affecting the Life Expectancy of an Individual) Life expectancy can be defined as the number of years the individual can expect to live. The health of the populations can be accurately measured with the life expectancy of an individual. Using the life expectancy rate, the life expectancy of an individual can be easily summarized using the current mortality rates. Life expectancy is affected by several different factors such as socio-economic (income , employment, education and economic well-being), the ability of the people to access quality healthcare, healthcare quality, nutrition, environmental factors such as quality of air, availability of clean water, adequate sanitization and genetic factors. The paper seeks to evaluate the different factors that can affect the life expectancy of an individual. A central focus The paper will help resolve in identifying the key variables that are expected to affect the life expectancy of an individual. There have been several types of research about the factors affecting the life expectancy of an individual. Chan & Kamala Devi (2012) in their research found that life expectancy is greatly influenced by the availability of healthcare resources and higher levels of socio-economic advantages. The demographic changes increase life expectancy by way of healthcare resources. Their research found evidence of improving healthcare coverage to improve life expectancy. This evidence was further supported by research done by Hosokawa et.al. (2020) where it was found that life expectancy is significantly affected by the number of doctors and therapists, support clinics for home healthcare and home visits treatment. Blinova et al. ( 2020) in their study about the reasons for a low life expectancy of the rule population in Russia found statistically significant social, behavioral and economic factors affecting the life expectancy of an individual. This topic is of interest to me and other researchers because the identification of factors affecting life expectancy can help in improving the average life of an individual. For example, based on past researches, it is clear that improving the health condition of the given regions of the country can dramatically improve the individual's life expectancy. Further, shifts in life expectancy are often used to describe the trends in mortality. Life expectancy being able to predict how populations will age has a significant influence on the planning and provisions of services and support provided by the government. The rationale for the project This project will help in identifying the different factors such as income, adult mortality, infant deaths, presence of disease such as HIV/AIDS, diphtheria, polio, schooling, GDP and alcohol consumption that can affect the life expectancy of an individual. The paper is unique in the sense it takes into account different social, economic, and environmental factors to predict the life expectancy of an individual. Further, it takes into account the impact of both countries and the trend in analyzing the individual's life expectancy. A panel data regression model will be developed for analyzing the cross-sectional and longitudinal data. This project considers the impact of both time and cross-section which makes it more efficient than time-series data. A statement of Methodology The dataset for the study will be downloaded from the third-party Kaggle website. However, the primary source of the dataset is the World Health Organization that keeps track of the health status as well as different factors for all countries. The dataset used is secondary as it is directly downloaded from the Kaggle website. The type of research includes secondary source research. References Department of Health. (n.d.). Tier 1—Life expectancy and wellbeing—1.19 Life expectancy at birth. Www1.Health.gov.au. https://www1.health.gov.au/internet/publications/publishing.nsf/Content/oatsih-hpf-2012toc~tier1~life-exp-wellb~119#:~:text=Life%20expectancy%20is%20affected%20by Blinova, T., Bylina, S., & Rusanovskiy, V. (2020). FACTORS AFFECTING THE LIFE EXPECTANCY AT BIRTH OF THE RURAL POPULATION IN RUSSIA. PONTE International Scientific Research Journal, 76(1). https://doi.org/10.21506/j.ponte.2020.1.2 Chan, M. F., & Kamala Devi, M. (2012). Factors Affecting Life Expectancy. Asia Pacific Journal of Public Health, 27(2), 136–146. https://doi.org/10.1177/1010539512454163 Hosokawa, R., Ojima, T., Myojin, T., Aida, J., Kondo, K., & Kondo, N. (2020). Associations between Healthcare Resources and Healthy Life Expectancy: A Descriptive Study across Secondary Medical Areas in Japan. International Journal of Environmental Research and Public Health, 17(17), 6301. doi:10.3390/ijerph17176301 panel data regression model analysis Assessing the use of panel data regression model The first model to test was a multiple linear regression model. The assumptions for the use of this mode were tested and the results interpreted. The tests included test for normality, test for, independence, linearity, and equal variance assumption. In the testing for normality assumption, the quantile plot of the residuals showed a an approximately linear plot. The normal quantile plot of residuals did not form linear diagonal line. This showed that the normality assumption was not satisfied by the data. While testing for the random assumption, the assumption could not be verified from the data because all the countries were sampled. For the independence assumption, the data collected are independent because they are collected from different countries for different measurements. Because of the independence sampling of data, independence assumption is satisfied. The assumption for equal variance was not satisfied because the plot of residuals against the predicted life expectancy is not spread uniformly around the line. Since some conditions were not met for using a multiple linear regression, the panel data regression model was the most preferred. Because of these violations, the multiple linear regression model was dropped in favor of the panel data regression model. Plot of residuals against predicted. Normal quantile plot The panel regression model is preferred because it takes into account both the random and fixed effect and minimizes bias. It can therefore create a better trend as compared to a multiple linear regression. Also, since the conditions for linear regression were violated, the panel regression that used an OSL model could not be used. Pane Data Regression Model The panel model appeared to be the best option for this study. The panel model is useful in creating trend in data that has been collected over time. Testing assumptions for the model, the quantile plot of residuals formed a relatively linear diagonal line. This showed that the normality assumption was satisfied by the residual plot. This was also displayed by the histogram of the data. The histogram of the data appeared symmetrical. While testing for the random assumption, the assumption could not be verified from the data because all the countries were sampled. For the independence Assumption, the data collected are independent because they are collected from different countries for different measurements. Because of the independence sampling of data, independence assumption is satisfied. The assumption for equal variance was not satisfied because the plot of residuals against the predicted life expectancy is not spread uniformly around the line. Even though some conditions were not met, the model satisfied most conditions and can therefore be used. A plot of Residuals against predicted A normal quantile plot and a histogram of residuals. A panel data has both a cross-section and a time-series dimension. In this data, all timeseries observations are observed during the whole time period. Xit,i=1,2,….n,t=1,.2…T T is generally small. The standard static panel data model with i=1,…N ,t=1,…T is Yit =β0 +x’itβ +€it. Here, xit is a K-dimensional vector of explanatory variables without a constant term β0 is the intercept which is not dependent on i and t. β a (K×1) vector, the slopes are independent of I and t. €it. , the error varies over I and t. To successfully use this model technique, there are a few assumptions that are considered. The model can be designed to exhibits a constant coefficient for slope but has a changing intercept over time. The model can also be designed to have both the slope and intercept coefficients vary over time. Using these assumptions, models of fixed effect estimation and random effects estimation are created. These two models enable the panel regression model to take into account both time and trend factors in prediction. In this project, I used a fixed-effects model for the analysis because the fixed effects model is less biased and is not affected highly by the assumptions. Reference Cao, J. (2015, March 24). Linear Mixed Models with JMP® Pro: One Face, Many Flavours. Jmp.Com. https://community.jmp.com/t5/Discovery-Summit-Europe-2015/LinearMixed-Models-With-JMP-Pro-One-Face-Many-Flavours/tap/22333#:~:text=JMP%20Pro%2011%20has%20added,to%20its%20Fit%20Model%20p latform.&text=Linear%20mixed%20models%20are%20a,%3DX%CE%B2%2BZ%CE% B3%2B%CE%B5 . JMP. (n.d.). How do I build Repeated Measures and Panel Data Models? https://community.jmp.com/t5/Short-Videos/How-do-I-build-Repeated-Measures-andPanel-Data-Models/ta-p/323564 JMP Help. (2021, May 5). Retrieved May 07, 2021, from https://www.jmp.com/support/help/en/16.0/index.shtml#page/jmp/mixed-models-andrandom-effect-models.shtml#ww739671 Vella, F., & Verbeek, M. (1998). Whose Wages do Unions Raise? A Dynamic Model of Unionism and Wage Rate Determination for Young Men. Journal of Applied Econometrics, 13(2), 163-183. Retrieved May 7, 2021, from https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.18.3907&rep=rep1&type=pdf