Measuring a Smile (R, Tableau project)

(U.S Happiness Report Analysis and suggestion for business)

Clara Lee
8 min readNov 19, 2022


What is the project’s context?

· Origin Curiosities

While getting through Covid-19 pandemic, one of the phenomena observed in the U.S. was the great resignation. Especially, so-called “essential workers” found themselves under-compensated and overworked by their employers. They didn’t want to take significant risks for shallow praise for keeping essential services available and left for other jobs with better pay and benefits. Many companies shifted toward remote work for office employees and those employees could save more money and time for their hobbies, family, or pet care. What do they tell us? Could it be all about pursuing a happiness? How to tell if people are happy or not? Can a happiness be measured?

Since 2012, a World Happiness Report have been releases every three years. The report is based on the survey asking approximately 1000 residents(age 15 and older) per country how they feel about their lives on a scale from zero to ten. Several other factors were measured that might affect the happiness score: household income, donation status, social relationships, and perception of government etc. Among more than 150 countries, the U.S ranks at the 16th happiest country in the world in 2022, and that’s a three-spot gain from last year, when was 19th. What’s the effect of higher happiness rank? Does economic situation get better? What made the happiness score go up?

PowerPoint by Clara Lee

According to the research, the happiest healthiest workplaces enjoy a 13% higher level of productivity and performance, their employees stay in their jobs for longer and the workplace becomes an attractant for other top talent. This effect of emotions on behavior might lead us to those questions above, and this project will be a part of the help to understand of how the ‘happiness’ could a benefit of our society.

· Circumstances

The project is going to be undertaken online using open data source, which was collected by Gallop World Poll. The data is downloaded as csv file from the Kaggle and cleaned, processed in R, and visualized using Tableau.

· Constraints

Due to limited situation ,the live presentation would not be available.

· Consumption

The audiences will consume the result through a power point form. It contains both data visualization and the explanation and details of findings to help understand what the charts or tables want to tell.

· Deliverables

I will create a powerpoint with a successful outcome that all the key information included.

· Resources

World Happiness Report ( )

Load the data

The data has total 1953 observations and 12 variables. There are two dimensions, Country name and year, and nine metrices, Happiness score,Economy, Social support, Life expectancy, Freedom, Generosity, Corruption, Positive affect, and Negative affect.

Happiness Report Data from PowerPoint by ClaraLee


o Happiness score :current life scale 0 (worst) to 10 (best)

o Economy : GDP per capita

o Social Support : if there’s person that you can count on ; no:0, yes:1

o Life Expectancy : life expectancy (age) at birth

o Freedom : make life choice; satisfied:1, dissatisfied:0

o Generosity :if you donate money to a charity in the past month: no:0, yes:1

o Corruption : perception of corruption on government and business scale 0 to 1

Clean the data

I already changed the name of variables in Excel and loaded in R. I don’t want to use Positive effect and Negative effect matrices here, so I’ll drop them and create another dataframe ‘smile’ and then, change the data type for metrices from character to numeric for analysis.

The distribution and possible outliers were checked by the histogram. Since Social support, Freedom, Corruption, and Generosity are binary data, I checked Happiness score, Economy, and Life expectancy.

Happiness score shows nearly normal distribution. The center is between 5 and 5.5. Economy plot shows bimodal around 9.3 and 10.8. There seems to have a slight skewness to the left. The Life expectancy plot is fairly left skewed , but it is unimodal around age 66. There are possible outliers under age 40.

Then, I created the subset for United States data, and checked the summary.

There is not many null data existing. One in social support , and the other is in Generosity. They would not affect the result and I ignored them for now.

There is one possible outlier in Happiness score but the happiness score would be dependent variable for the analysis, so I decided not to remove it to find out why it has a higher value.

The scatter plots were created to see the correlation between the variables for world data. The plots show that they have a positive linear correlation each other. It indicates that when happiness score increases, economy and life expectancy increase.

World happiness correlation

How about the U.S? Are variables for U.S correlated each other?

U.S happiness correlation

In the U.S plots, there is a weak negative correlation between Happiness score and Economy. This is somewhat different from the result of world data.

I created visualizations in Tableau and made slides on PowerPoint.

This is to create a data story .

from PowerPoint by Clara Lee

This is the map chart showing world happiness score. As we can see, North America, South America, Europe and Australia and Oceania tend to have higher happiness score while Africa, Middle east, and Asia have lower happiness score. The U.S has the average happiness score of 7.09 between year of 2005 and 2020. It is higher than average world happiness score of 5.52.

from PowerPoint by Clara Lee

I looked into the trend of happiness score and compared U.S with World data. Since 2005 data is missing in U.S, let’s compare from year of 2006. (The big drop from 2005 to 2006 in world data looks interesting though) The score in the U.S decreased between 2007 and2016, and started bouncing back while the score in World has been showing increasing trend.

PowerPoint by Clara Lee

All top 20 countries from 2005 to 2020 have higher ranks in Economy. However, the rank does not match the rank of happiness. What does it tell us? According to a large research by the Gallup World Poll, people are happiest when they make between $60,000 and $75,000 a year. It means there could be other factors make you feel happy once you can make a certain amount of income. As we can see from the graph, the U.S ranked at 5th in average Economy but ranked at 16 in average Happiness score. And Happiness score and Economy have a negative linear relationship. The equation says that one unit increase in happiness score is associated with -0.1612 decrease in Economy. How people in the U.S feel themselves happy is more affected by something else , and not greatly by money.

PowerPoint by Clara Lee

I was not surprised to see the result of world data how Happiness score and Life expectancy correlated. They show almost same trend. Meanwhile, their relationship is not significant in the U.S. Living a healthy long life do not affect much people in the U.S/

“Then, What Factors Follow the Trend of Happiness Score in the U.S?”

Let’s take a look at this chart.

From Tableau Dashboard by Clara Lee

Generosity, Freedom, Social Support show similar trend to Happiness score, even though Generosity dropped after 2019 (the pandemic hit). Corruption overall shows opposite trend from Happiness score.

How much correlation do Generosity, Freedom , and Social support have with Happiness score?

From Tableau Dashboard by Clara Lee

It has turned out that Social support is the most correlated factor with R-square 0.41 and p-value is less than 0.05, which is significant. Generosity also has a moderate positive correlation with R-square of 0.31 and p-value less than 0.05. Freedom and Corruption don’t have significant correlation with Happiness score.


from PowerPoint by Clara Lee

Happiness can be measured and we can increase it by ‘Social support’. I didn’t include ‘Generosity’ in the conclusion even though it is still significant factor to predict Happiness score. This is because I wanted to approach the data to find a way to increase the happiness score from business perspective, and the most impactful factor is ‘Social support’.

Then, how can we increase happiness at WORK ? How to create a happy workplace? What can company and leaders do for employees?

Authenticity at work is showing up as your whole self. However, showing authenticity at work doesn’t happen in an instance. It actually requires lots of work and time because it is possible when employees feel psychologically safe, inclusion, trust, and belonging. That is the true feeling supported at work and would be a big factor to be happy at work.

The inclusive and authentic leadership should come here. As suggested in the slide above,

  • Lead with empathy
  • Show care for the whole person
  • Listen to feedback
  • Create safe spaces for difficult conversation
  • Lead with inclusivity and belonging
  • Support personal and professional development
  • Provide access to personalized coaching
  • Invest in your leaders

These can empower the growth of employees and promote their happiness at work. I believe the outcome from those efforts could give business and society a great benefit.



Clara Lee

Data Story Teller | Data is my passion | Connect with me ➡️