SCATTER PLOTS AND CORRELATION

[Link Map]
begin comments text sound
line

Scatter plots show simple relationships between variables and are only useful if at least one of the variables is continuous. The result is more useful if both variables are continuous as shown using the weather data below. The second plot has been improved by formatting the y-axis so that the lower limit is 40.
  
The correlation between pairs of variables can be calculated by: Tools > Data Analysis... > Correlation. Selecting temp, rad and humidity as the input ranges creates the following correlation table: Notice that the correlation coefficient is positive for the first pair and negative for the second. Correlation measures the strength of the relationship between two variables, always lies between -1 and +1 and can be interpreted as follows:
  • Close to -1
    		•  ==> strong negative relationship (the plot should have a reasonably straight line from top left to bottom right)
  • Between -1 and 0
    		• ==> negative relationship
  • About 0
    		• ==> little relationship
  • Between 0 and 1
    		•  ==> positive relationship
  • Close to 1
    		•  ==> strong positive relationship (the plot should have a reasonably straight line heading from bottom left to top right)

    The order in which you specify the variables is not important. The correlation coefficient can give misleading answers with more complicated relationships since it only measures the strength of straight line relationships. Also, correlation does not imply cousation.

    line
    Basic statistics in Excel   23.2.99   Page: 10 of 25