To install StudyMoose App tap and then “Add to Home Screen”
Save to my list
Remove from my list
For my class project, I used the data from the 2019 7th World Happiness Report. Each year the report focuses on some themes of well being and happiness. This year the report was on happiness and the community and looked at the link between government actions and happiness, happiness and community, and happiness and technology. Their study focused on six explanatory factors: social support, healthy life expectancy, GDP per capita, generosity, freedom to make life choices, and freedom from corruption. I chose this topic as, during the midst of the COVID019 pandemic, mindfulness and wellness have been common themes in the media.
In this study, the Cantril Ladder was used to measure overall well-being.
The Cantil Ladder is often compared to a 10 rung step ladder that we may have in our tool shed, with the first rung being a score of 1 and the 10th rung rated as a score of 10. Features included in the study: GDP per capita, social support, healthy life expectancy, freedom, generosity, and absence of corruption.
Verified writer
Proficient in: Data Analysis
4.8 (309)
“ Writer-marian did a very good job with my paper, she got straight to the point, she made it clear and organized ”
The study uses the variables to explain the variation of happiness across countries. For my project, I focused on the impact of social support, healthy life expectancy, GDP, and generosity on happiness. I also analyzed the factors contributing to a healthy life expectancy.
Two questions I have are:
My data was rank-ordered from 1 to 156, so I found no outliers. In my initial analysis, I only considered four features: social support, generosity, GDP per capita, and healthy life expectancy. 12 null values were detected, and I chose to delete the rows with null values.
The table below shows what I found during my initial look at the data.
Using the features that I had selected initially, social support, generosity GDP per capita, and healthy life expectancy, I plotted the data for a visual representation of the relationships. Using KNIME for the analysis tool, I found that three of the four features had similar scatter plots so I could not tell which variable was more dominant. I then turned to a linear regression model, a formal supervised machine learning technique. R2 is a statistical measure of fit that indicates how much variation of a dependent variable is explained by the independent variable(s) in a regression model. As shown on the next page, the inclusion of all features in the original study provided the highest R2, meaning that the inclusion of all features was needed in the model to optimize the likelihood of happiness. R2 never reached 1, meaning that some other factor was missing from the model or that despite the inclusion of all the variables, there was no guarantee that people would be happy despite the most optimal conditions.
The scatterplot (left) of the four features shows social support is the most influential feature of happiness. This combination of features yielded an adjusted R2 of 0.7816.
I experimented with the features used in the study to find the “optimal” combination to find happiness. After trial-and-error, I included the six features in the original study plus what the author referred to as the positive effect and the negative effect. Per the authors of the original study, “Positive effect comprises the average frequency of happiness, laughter, and enjoyment on the previous day, and negative affect comprises the average frequency of worry, sadness, and anger on the previous day. The affect measures thus lie between 0 and 1.” The negative effect lowers the overall score by .08 but considered it necessary as we were including the positive effect. The additional features raised the R2 value by .0328. See below.
The scatterplot (left) of the original six features as well as the positive and negative affect were used to assess how the features contribute to happiness. This combination of features yielded an adjusted R2 of 0.816.
During the pandemic, many people are living in social near-isolation if they are without some tools such as Zoom and Facetime. My initial thought was that the lack of social support would have an adverse impact on a healthy life expectancy. The data below shows that GDP per capita has the greatest influence on a healthy life expectancy. In this case, we have an adjusted R2 of 0.7289 with social support, generosity, and GDP per capita factored into the equation.
The scatterplot (left) of the original six features as well as the positive and negative affect were used t to assess how the features contribute to happiness. This combination of features yielded an adjusted R2 of 0.816.
Again, trying to find the optimal combination of features to predict a healthy life expectancy, I added freedom and lack of government corruption to the equation and raised the adjusted R2 to 0.7708. GDP per capita continues to be a leading feature in determining the likelihood of a healthy life expectancy.
The scatterplot (left) of the original six features as well as the positive and negative affect were used t to assess how the features contribute to happiness. This combination of features yielded an adjusted R2 of 0.816.
I chose to use cluster analysis for unsupervised machine learning. For my first question, what features most influence happiness, I would include in my analysis three of the four features I identified in my initial list of features: social support, healthy life expectancy, and GDP per capita. I would cluster common happiness indicator scores and look to see what scores assigned to the features and look for commonalities. For clusters with a similar happiness score, what range of scores are associated closely with each of the features?
For my second question, what features most influence a healthy life expectancy, I would perform a similar analysis as mentioned above. I would group (cluster) similar healthy life expectancy scores and look at the scores assigned to the features GDP per capita, social support, and generosity.
From a review of the scatter plots alone, it was difficult to tell which factors had the strongest impact on happiness and healthy life expectancy. Regression analysis assisted in assessing the best combination of features for happiness as well as a healthy life expectancy. All features included in the original study provided the best indicator of happiness
Happiness and Life Satisfaction - Our World in Data. (2020, Sep 02). Retrieved from https://studymoose.com/happiness-and-life-satisfaction-our-world-in-data-essay
👋 Hi! I’m your smart assistant Amy!
Don’t know where to start? Type your requirements and I’ll connect you to an academic expert within 3 minutes.
get help with your assignment
