Hi,

The reasoning for your differing answers is due to your procedure in calculating the Pearson coefficients.

Without directly telling you the answer, I would urge you to reread this part of the project brief, I have highlighted the important parts:

"The method is: for each predictor, for each country, compute the correlation (also known as the Pearson correlation) between the predictor and the associated Cantril Ladder score. The mean correlation for a given predictor is then computed across all the countries".

Hint for python (if you are using pandas) you will need to make at least one df.groupby() calls.

Hope that helps :)