First Let’s take a peak of our dataset:

happiness_df %>% head() %>% knitr::kable()
country_name regional_indicator year ladder_score logged_gdp_per_capita social_support healthy_life_expectancy freedom_to_make_life_choices generosity perceptions_of_corruption
Afghanistan South Asia 2021 2.523 7.695 0.463 52.493 0.382 -0.102 0.924
Afghanistan South Asia 2019 2.375 7.697 0.420 52.400 0.394 -0.108 0.924
Afghanistan South Asia 2018 2.694 7.692 0.508 52.600 0.374 -0.094 0.928
Afghanistan South Asia 2017 2.662 7.697 0.491 52.800 0.427 -0.121 0.954
Afghanistan South Asia 2016 4.220 7.697 0.559 53.000 0.523 0.042 0.793
Afghanistan South Asia 2015 3.983 7.702 0.529 53.200 0.389 0.080 0.881

The dataset has 1816 observations and 10 variables, Let’s take a deeper look into each variable:

EDA of each variable

Country name

happiness_df %>%
  group_by(country_name) %>%
  summarize(
    n_obs = n()
  ) %>%
  group_by(n_obs) %>%
  summarize(
    number_of_countries = n()
  ) %>%
  knitr::kable()
n_obs number_of_countries
1 6
3 2
4 6
5 3
6 2
7 4
8 3
9 6
10 8
11 4
12 7
13 12
14 31
15 31
16 24

There are 149 unique countries in this dataset. We can see that 6 countries appeared only 1 time, while 24 countries appeared 16 times. The countries that appeared only 1 time are China, Hong Kong, Kosovo, Maldives, North Cyprus, and Turkmenistan.

Regional indicator

There are 10 unique regions in this dataset. The regional indicators are South Asia, Central and Eastern Europe, Middle East and North Africa, Latin America and Caribbean, Commonwealth of Independent States, North America and ANZ, Western Europe, Sub-Saharan Africa, Southeast Asia, East Asia.

happiness_df %>%
  group_by(regional_indicator) %>%
  summarize(
    n_obs = n()
  ) %>%
  mutate(regional_indicator = fct_reorder(regional_indicator, n_obs)) %>%
  ggplot(aes(x = regional_indicator, y = n_obs)) +
  geom_col() +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5)) +
  labs(
    x = "regional indicator",
    y = "number of observations",
    title = "Number of observations for each region"
  )

East Asia has the lowest number of observations, while Sub-Saharan Africa has the highest.

Year

happiness_df %>%
  group_by(year) %>%
  summarize(
    n_obs = n()
  ) %>%
  ggplot(aes(x = factor(year), y = n_obs)) +
  geom_col() +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5)) +
  labs(
    x = "year",
    y = "number of observations",
    title = "Number of observations for each year"
  )

Year ranges from 2005 to 2021 in this dataset. The year with the most observation is 2021, and the year with the least observation is 2005, which only has 1 observation.

Ladder score

happiness_df %>%
  ggplot(aes(x = ladder_score)) +
  geom_histogram(bins = 20) +
  labs(title = "Distribution of ladder score")

Ladder score ranges from 2.375 to 7.971. The distribution is approximately symmetric with mean at 5.470168.

Since ladder score is the main outcome of our project, we should dive a little deeper by looking at the mean ladder score for each region and year.

happiness_df %>%
  group_by(regional_indicator) %>%
  summarize(
    mean_ladder_score = mean(ladder_score)
  ) %>%
  arrange(mean_ladder_score) %>%
  knitr::kable()
regional_indicator mean_ladder_score
Sub-Saharan Africa 4.317518
South Asia 4.512207
Commonwealth of Independent States 5.180830
Middle East and North Africa 5.271464
Southeast Asia 5.368957
Central and Eastern Europe 5.550598
East Asia 5.682980
Latin America and Caribbean 5.988366
Western Europe 6.871429
North America and ANZ 7.246328

Sub-Saharan Africa has the lowest mean ladder score across all years, while North America and ANZ has the highest.

happiness_df %>%
  group_by(year) %>%
  summarize(
    mean_ladder_score = mean(ladder_score)
  ) %>%
  knitr::kable()
year mean_ladder_score
2005 7.418000
2006 5.183946
2007 5.367533
2008 5.437252
2009 5.480366
2010 5.546394
2011 5.411216
2012 5.446638
2013 5.379185
2014 5.396512
2015 5.404903
2016 5.418459
2017 5.505721
2018 5.542738
2019 5.559349
2020 5.873580
2021 5.532839

Year 2005 has significantly higher ladder score than the others. However, because year 2005 only has one observation, the difference should not be considered significant. Ladder scores for all other years are around 5.5.

Logged GDP per capita

happiness_df %>%
  ggplot(aes(x = logged_gdp_per_capita)) +
  geom_histogram(bins = 15) +
  labs(title = "Distribution of logged GDP per capita")

Logged GDP per capita ranges from 6.635 to 11.648. The distribution is approximately symmetric with mean at 9.3522203.

Social support

happiness_df %>%
  ggplot(aes(x = social_support)) +
  geom_histogram(bins = 15) +
  labs(title = "Distribution of social support")

Social support ranges from 0.291 to 0.987. Since the distribution is skewed right, we will look at the median instead of the mean. The median of social support is 0.836.

Healthy life expectancy

happiness_df %>%
  ggplot(aes(x = healthy_life_expectancy)) +
  geom_histogram(bins = 15) +
  labs(title = "Distribution of healthy life expectancy")

Healthy life expectancy ranges from 32.3 to 77.1. The distribution is skewed right. The median of healthy life expectancy is 65.3015.

Freedom to make life choices

happiness_df %>%
  ggplot(aes(x = freedom_to_make_life_choices)) +
  geom_histogram(bins = 15) +
  labs(title = "Distribution of freedom to make life choices")

Freedom to make life choices ranges from 0.258 to 0.985. The distribution is skewed right. The median of freedom to make life choices is 0.766.

Generosity

happiness_df %>%
  ggplot(aes(x = generosity)) +
  geom_histogram(bins = 15) +
  labs(title = "Distribution of generosity")

Generosity ranges from -0.335 to 0.689. The distribution is approximately symmetric with mean at -0.0031773.

Perceptions of corruption

happiness_df %>%
  ggplot(aes(x = perceptions_of_corruption)) +
  geom_histogram(bins = 15) +
  labs(title = "Distribution of perceptions of corruption")

Perceptions of corruption ranges from 0.035 to 0.983. The distribution is skewed right. The median of perceptions of corruption is 0.804.

Summary

Finally, let’s take a look at the summary tables of mean and standard deviation of each variable.

happiness_df %>%
  select(ladder_score, logged_gdp_per_capita, social_support, healthy_life_expectancy, freedom_to_make_life_choices, generosity, perceptions_of_corruption) %>%
  tbl_summary(missing_text = "(Missing)",
              statistic = list(all_continuous() ~ "{mean} ({sd})"),
              label = list(ladder_score ~ "Ladder score",
                           logged_gdp_per_capita ~ "Logged gdp per capita",
                           social_support ~ "Social support",
                           healthy_life_expectancy ~ "Healthy life exp",
                           freedom_to_make_life_choices ~ "Freedom to make life choices",
                           generosity ~ "Generosity",
                           perceptions_of_corruption ~ "Perceiptions of corruption")) %>% 
  bold_labels() %>%
  italicize_levels()
Characteristic N = 1,8161
Ladder score 5.47 (1.13)
Logged gdp per capita 9.35 (1.15)
Social support 0.81 (0.12)
Healthy life exp 64 (8)
Freedom to make life choices 0.75 (0.14)
Generosity 0.00 (0.16)
Perceiptions of corruption 0.75 (0.19)
1 Mean (SD) 
happiness_df %>%
  select(regional_indicator, ladder_score, logged_gdp_per_capita, social_support, healthy_life_expectancy, freedom_to_make_life_choices, generosity, perceptions_of_corruption) %>%
  tbl_summary(by = "regional_indicator",
              missing_text = "(Missing)",
              statistic = list(all_continuous() ~ "{mean} ({sd})"),
              label = list(ladder_score ~ "Ladder score",
                           logged_gdp_per_capita ~ "Logged gdp per capita",
                           social_support ~ "Social support",
                           healthy_life_expectancy ~ "Healthy life exp",
                           freedom_to_make_life_choices ~ "Freedom to make life choices",
                           generosity ~ "Generosity",
                           perceptions_of_corruption ~ "Perceiptions of corruption")) %>% 
  bold_labels() %>%
  italicize_levels()
Characteristic Central and Eastern Europe, N = 2241 Commonwealth of Independent States, N = 1651 East Asia, N = 491 Latin America and Caribbean, N = 2921 Middle East and North Africa, N = 1381 North America and ANZ, N = 581 South Asia, N = 871 Southeast Asia, N = 1161 Sub-Saharan Africa, N = 4211 Western Europe, N = 2661
Ladder score 5.55 (0.63) 5.18 (0.62) 5.68 (0.51) 5.99 (0.76) 5.27 (0.96) 7.25 (0.18) 4.51 (0.69) 5.37 (0.75) 4.32 (0.64) 6.87 (0.69)
Logged gdp per capita 10.03 (0.36) 9.28 (0.66) 10.17 (0.64) 9.33 (0.62) 9.61 (0.76) 10.76 (0.14) 8.33 (0.55) 9.26 (0.96) 7.99 (0.76) 10.79 (0.28)
Social support 0.86 (0.08) 0.82 (0.11) 0.87 (0.06) 0.85 (0.07) 0.77 (0.10) 0.94 (0.02) 0.65 (0.13) 0.81 (0.07) 0.71 (0.11) 0.92 (0.05)
Healthy life exp 67 (2) 63 (2) 70 (6) 66 (5) 65 (4) 72 (2) 60 (5) 64 (5) 53 (5) 72 (1)
Freedom to make life choices 0.69 (0.14) 0.70 (0.13) 0.71 (0.09) 0.78 (0.11) 0.64 (0.13) 0.90 (0.04) 0.69 (0.17) 0.87 (0.07) 0.70 (0.11) 0.84 (0.12)
Generosity -0.10 (0.10) -0.12 (0.14) -0.04 (0.12) -0.03 (0.12) -0.06 (0.13) 0.23 (0.08) 0.08 (0.11) 0.20 (0.22) -0.01 (0.11) 0.06 (0.17)
Perceiptions of corruption 0.87 (0.09) 0.78 (0.16) 0.79 (0.11) 0.79 (0.09) 0.77 (0.13) 0.45 (0.16) 0.81 (0.07) 0.73 (0.25) 0.79 (0.13) 0.56 (0.24)
1 Mean (SD) 
happiness_df %>%
  select(year, ladder_score, logged_gdp_per_capita, social_support, healthy_life_expectancy, freedom_to_make_life_choices, generosity, perceptions_of_corruption) %>%
  filter(year != 2005) %>%
  tbl_summary(by = "year",
              missing_text = "(Missing)",
              statistic = list(all_continuous() ~ "{mean} ({sd})"),
              label = list(ladder_score ~ "Ladder score",
                           logged_gdp_per_capita ~ "Logged gdp per capita",
                           social_support ~ "Social support",
                           healthy_life_expectancy ~ "Healthy life exp",
                           freedom_to_make_life_choices ~ "Freedom to make life choices",
                           generosity ~ "Generosity",
                           perceptions_of_corruption ~ "Perceiptions of corruption")) %>% 
  bold_labels() %>%
  italicize_levels()
Characteristic 2006, N = 741 2007, N = 921 2008, N = 991 2009, N = 1011 2010, N = 1091 2011, N = 1251 2012, N = 1161 2013, N = 1191 2014, N = 1231 2015, N = 1241 2016, N = 1221 2017, N = 1291 2018, N = 1261 2019, N = 1261 2020, N = 811 2021, N = 1491
Ladder score 5.18 (1.11) 5.37 (1.02) 5.44 (1.20) 5.48 (1.06) 5.55 (1.15) 5.41 (1.10) 5.45 (1.16) 5.38 (1.20) 5.40 (1.21) 5.40 (1.15) 5.42 (1.15) 5.51 (1.13) 5.54 (1.09) 5.56 (1.13) 5.87 (1.00) 5.53 (1.07)
Logged gdp per capita 9.05 (1.22) 9.17 (1.11) 9.16 (1.25) 9.29 (1.14) 9.42 (1.11) 9.25 (1.16) 9.34 (1.10) 9.34 (1.11) 9.36 (1.15) 9.36 (1.17) 9.40 (1.16) 9.39 (1.16) 9.40 (1.18) 9.42 (1.15) 9.75 (0.95) 9.43 (1.16)
Social support 0.83 (0.12) 0.81 (0.11) 0.78 (0.15) 0.81 (0.11) 0.83 (0.10) 0.81 (0.13) 0.81 (0.12) 0.81 (0.11) 0.81 (0.12) 0.80 (0.13) 0.82 (0.12) 0.81 (0.12) 0.81 (0.12) 0.81 (0.12) 0.84 (0.11) 0.81 (0.11)
Healthy life exp 60 (9) 61 (8) 61 (9) 63 (8) 63 (8) 62 (8) 63 (7) 63 (7) 64 (7) 64 (7) 64 (7) 64 (7) 65 (7) 65 (7) 67 (6) 65 (7)
Freedom to make life choices 0.73 (0.13) 0.69 (0.15) 0.68 (0.16) 0.68 (0.15) 0.71 (0.15) 0.73 (0.15) 0.71 (0.15) 0.73 (0.13) 0.74 (0.14) 0.75 (0.13) 0.76 (0.13) 0.78 (0.13) 0.78 (0.12) 0.79 (0.12) 0.82 (0.09) 0.79 (0.11)
Generosity 0.00 (0.17) 0.01 (0.17) 0.02 (0.17) -0.02 (0.16) 0.00 (0.16) -0.02 (0.17) 0.00 (0.17) 0.00 (0.17) 0.02 (0.16) 0.02 (0.17) 0.00 (0.16) -0.01 (0.16) -0.02 (0.16) -0.02 (0.15) 0.00 (0.14) -0.02 (0.15)
Perceiptions of corruption 0.76 (0.18) 0.80 (0.17) 0.77 (0.19) 0.78 (0.18) 0.77 (0.18) 0.76 (0.18) 0.77 (0.18) 0.76 (0.18) 0.74 (0.19) 0.74 (0.20) 0.75 (0.19) 0.72 (0.18) 0.73 (0.19) 0.72 (0.19) 0.71 (0.20) 0.73 (0.18)
1 Mean (SD)

Note: Since year 2005 only has one observation, it is removed from this summary table.