data. Its value here is +0.6 which indicates a limited measure of agreement
between Albert and Zac. If we treat the pairs of ranks as points we can plot them
on a graph to obtain a visual representation of how closely the two judges agree.
The formula for this correlation coefficient was developed in 1904 by the
psychologist Charles Spearman who, like Pearson, was influenced by Francis
Galton.
Measuring agreement between two judgesRegression lines
Are you shorter or taller than both your parents or do you fall between their
heights? If we were all taller than our parents, and this happened at each
generation, then one day the population might be composed of ten-footers and
upwards, and surely this cannot be. If we were all shorter than our parents then
the population would gradually diminish in height and this is equally unlikely.
The truth lies elsewhere.
Francis Galton conducted experiments in the 1880s in which he compared the
heights of mature young adults with the heights of their parents. For each value
of the x variable measuring parents’ height (actually combining height of mother
and father into a ‘mid-parent’ height) he observed the heights of their offspring.
We are talking about a practical scientist here, so out came the pencils and sheets
of paper divided into squares on which he plotted the data. For 205 mid-parents
and 928 offspring he found the average height of both sets to be 68¼ inches or