Disc 1

The r coefficient for the quoted data is not calculated, but I would assume that it is somewhere close to 0.5-0.6, as there is a clear positive correlation, but still the points are not situated perfectly along the hypothetical trend line.  Good observation! What will be the nature of points on a scatter diagram when the correlation is close to zero?

Original Question:

Provide a link for a picture of a scatter plot where the r value is not given. Estimate what you think the r value would be and explain why.

We answered:

In answer to this discussion, I have chosen to use a scatter plot from Washington Post, inspired by the recent successes of the Leicester City football team from the English Premier League (Washington Post, 2016). In essence, in reflects the visual representation of the relationship between the wage bills of the football players and the total points which they eventually score. The r coefficient for the quoted data is not calculated, but I would assume that it is somewhere close to 0.5-0.6, as there is a clear positive correlation, but still the points are not situated perfectly along the hypothetical trend line.
Therefore, it can be concluded that judging from the sample data, higher levels of wage bills are on average associated with more points scored (of this tendency, Leicester City is the outlier to some extent). Talking about causation component, it cannot be readily derived from this scatter plot: it can be that the players which are more talented end up having higher wages or, alternatively, that the high wages motivate the players to train more and eventually succeed. Furthermore, it can be the case that both of these effects are valid, or that some third unobserved variable causes this tendency.

Leave a Reply

Your email address will not be published. Required fields are marked *