Notes for comment questions

When exam questions ask you to comment on something, there is a standard, small, list of things that they are looking for. Make some notes about these as we go along.

Histograms:

• Are good for continuous data, which is then grouped into classes
• the vertical axis is the frequency density, frequency divided by class width
• The class width goes to half way between the end of this class and the beginning of the next, if the class boundaries aren't the same.

Scaling rules:

• mean(ax)=a.mean(x) - if the points are twice as big, so is the average
• stdev(ax)=a.stdev(x) - if the points are twice as big, they are twice as spread out.
• mean(x+a)=mean(x)+a - if you move all the points further along the axis, the mean is moved too
• stdev(x+a)=stdev(x) - if you move all the points further along the axis, they are still as spread as they were

box plots

• Show quartiles and median
• Show outliers
• Show skew
• Are good for comparing two datasets