Top Page | Upper Page | Contents | About This Site | JAPANESE

Scatter Plot and Correlation

Positive and negative correlation

If x-y data plot seems to be a line, it is called "there is correlation". Or " correlation is strong".

Correlation coefficient (R) is an index to measure correlation. If this index is +1 or -1, the plot has perfect linearity. If it is near 0, plots are scattered. This index is a kind of unevenness .

If it is near +1 or -1, this case is called "correlation". The case of +1 is called "Positive correlation". The case of -1 is called "Negative correlation". "Near" is not defined. We need to judge by studying the data.

NEXT Faults When We Find Correlation