Research Question

Question: Do countries where people help strangers more tend to be happier?

  • Response (Y): Happiness rank (Cantril Ladder rank)
  • Predictor (X): “Helped a stranger” rank (Benevolent act)
  • Data source: World Happiness Report 2025, Table 2.2 Country rankings for six measures of benevolence

A Happiness rank of 1 means the country is the most happy (highest average life evaluation) where a larger rank like 50 or 100 means the level of unhappiness.

Benevolent acts are behaviors like donating money, volunteering, or helping a stranger. A rank of 1 means that country had one of the highest rates of people helping strangers and vice versa.

Data (Top 34 countries from Table 2.2)

First 6 rows (key columns only)
country ladder_rank helped_rank happy_score helped_score
Finland 1 96 147 52
Denmark 2 76 146 72
Iceland 3 125 145 23
Sweden 4 90 144 58
Netherlands 5 134 143 14
Costa Rica 6 36 142 112

Quick Summary

Summary of the Statistics
happy_score helped_score donated_score volunteered_score
Min. :114.0 Min. : 2.00 Min. : 23.0 Min. : 3.00
1st Qu.:122.2 1st Qu.: 24.75 1st Qu.:106.2 1st Qu.: 61.75
Median :130.5 Median : 63.00 Median :119.5 Median : 88.00
Mean :130.5 Mean : 63.47 Mean :110.1 Mean : 82.06
3rd Qu.:138.8 3rd Qu.: 97.50 3rd Qu.:132.8 3rd Qu.:109.25
Max. :147.0 Max. :136.00 Max. :144.0 Max. :133.00

Helping a Stranger vs Happiness (with Regression Line)

ggplot(df, aes(x = helped_score, y = happy_score)) +
  geom_point() + geom_smooth(method = "lm", se = FALSE) +
  labs(x = "Helping a Stranger (higher score = more helping)",
       y = "Happiness (higher score = happier)")

The Model

Regression Coefficients
Estimate Std. Error t value Pr(>|t|)
(Intercept) 133.897 3.191 41.960 0.000
helped_score -0.054 0.043 -1.256 0.218

We use a simple linear regression:

\[ Y_i = \beta_0 + \beta_1 X_i + \varepsilon_i \] Where:

  • \(Y_i\) = happiness score (higher = better happiness rank)
  • \(X_i\) = helping-a-stranger score (higher = better helping rank)
  • \(\varepsilon_i\) = error term

Residual Plot

Hypothesis Test

To test whether helping is related to happiness:

\[ H_0: \beta_1 = 0 \quad \text{vs} \quad H_a: \beta_1 \ne 0 \]

Test statistic:

\[ t = \frac{\hat{\beta}_1 - 0}{SE(\hat{\beta}_1)} \]

If the p-value for \(\hat{\beta}_1\) is small (ex: < 0.05), we have evidence of a linear relationship.

3D Visualization: Helping, Donating, and Happiness (Top 34 Countries)

Each point is a country (Top 34 by happiness rank).
Axes are scores (higher = better rank). Color shows happiness score.

Conclusion + Limitations

What I found (Top 34 countries):

  • The fitted line is slightly negative.
  • Estimated slope: \(\hat{\beta}_1 =\) -0.054
    p-value = 0.218
  • Since the p-value is greater than 0.05, the evidence suggests there isn’t a linear relationship between helping and happiness unfortunately.

Limitations:

  • These are rank-based scores, not raw percentages of helping/donating.
  • Only Top 34 countries were used (small sample).
  • Country-level data shows association, not causation (many other factors affect happiness, there is still hope).

References