Mchikky/R-Analysis-of-the-African-World-Development-Indicators-data

Relationship between literacy rate and HIV prevalence # R project - Analysis of the African World Development Indicators data Relationship between literacy rate and HIV prevalence ## Objectives: To understand how literacy rate affects the prevalence of HIV among the adult population. ## Mission: - Identifies gaps and major concerns with the data and the methodology behind the data collection, as well as affirms quality methodological approaches - Presents descriptive statistics. - Explains decisions about an appropriate model(s). - Present model results in an engaging way. - Interprets the results in substantive terms ## Gaps: The data frame has 54 observations and 33 variables with various missing data points and blanks represented as NA and dots respectively. I am not sure of the methodology used in collecting this data, however, during data entry, the data type for antiretroviral therapy coverage (% of people living with HIV) was coded as “string” instead of “numeric”. Using the R studio data analysis software, I tried changing the column to numeric using “as.numeric” but it did not respond to the command, not sure what the problem was. Then I replaced “NA”s using zero. I checked for duplicates using “sum(duplicated()” command and there are no duplicates. I decided to work with 54 observations and 10 variables to test the hypothesis that literacy rate will likely affect the HIV prevalence among the adult population from 15-49 years. I looked at the variables such as life_expectancy, literacy, HIV_15_49, antiretroviral_HIV, population, urbanpop, ruralpop, and pop_growth. I checked for the Sub Saharan African countries with the highest prevalence rate of HIV which shows that Eswatini is topping the list with 28%. It seems that the HIV prevalence rate for non Sub Saharan countries such as Morocco, Libya, Egypt, and Algeria were not captured. (See the figure below) ## Descriptive Statistics: Using R studio, I installed certain packages including: “foreign”, “tidyverse”, ”pastecs”, “psych”, “ggplot2”, “corrplot”, and …