Data Analysis Tools: WEEK 3

Pearson Correlation

Assignment: Generating a Correlation Coefficient

Source: GapMinder
Variables used: incomeperperson, femaleemployrate, suicideper100TH (each of these variables is quantitative)

*incomeperperson – Gross Domestic Product (GDP) per capita
*femaleemployrate – Female Employment rate -Percentage of female population, age above 15, that has been employed during the given year.
*suicideper100TH - Suicide per 100th - Mortality due to self-inflicted injury, per 100 000 standard population, age adjusted

Introduction:

In this assignment, I decided to analyze the correlation in each pair of  two quantitative variables from the followings: GDP (per capita), Female Employment Rate and Suicide per 100th.

In my first attempt, I observed all the countries from the GapMinder set and it appeared that there is no significant relationship between any of the chosen variables.

After looking at the bivariate graphs, I noticed that in countries with the lowest GDP there is a much bigger spread of the results than in countries with higher GDP. Therefore, I decided to repeat the analysis and test the correlations for countries with GDP higher than 5000$ per capita:

CODE:

OUTPUT:

Interpretation:

GDP vs. Female Employment Rate

In countries with GDP higher than 5000$ per capita, the relationship between GDP and Female Employment Rate is significant. The p value is 0.0001. The Correlation Coefficient R is 0.48080 which means that there´s a positive association between the variables – the higher GDP, the higher Female Employment Rate.

R² equals 0.23116864, indicating that if we know GDP (explanatory variable) we can predict 23% of the variability we will see in the Female Employment Rate (response variable).

GDP vs. Suicide per 100th

In countries with GDP higher than 5000$ per capita (similarly to all other countries), the relationship between GDP and Suicide per 100th is NOT significant as the p value is equal to 0.8545 and the Correlation Coefficient R is 0.02440.

Female Employment Rate vs. Suicide per 100th

Interestingly, in countries with GDP higher than 5000$, there is a statistically significant relationship between Female Employment Rate and Suicide per 100th. The p value is equal to 0.0405. This relatively high p value (almost 0.05) indicates that the correlation between the two variables is not very strong.

The Correlation Coefficient R is 0.26759 so the association between the variables is positive, i.e. the Suicide per 100th increases with the increase of Female Employment rate.

R² equals 0.0716 which means that if we know the Female Employment Rate (explanatory variable) we can predict 7% of the variability we will see in the Suicide per 100th (response variable).