# Graphical representation of weekly spending

What is the correct graphical representation for these data of weekly spending (in dollars) on soft drinks for 20 people:

12,13,17,21,24,24,26,27,27,30,32,35,37,38,41,43,44,46,53,58


I want to separate the data into 5 bins:

10-20(f=3),    20-30(f=6),    30-40(f=5),    40-50(f=4),    50-60(f=2)


and I title axis: "Weekly spending (dollar amount)".

Which one would be more appropriate for y axis? (For which I put f=3,6,5,4,2.)

1. Frequency per \$1 spent 2. Frequency per \$10 spent
• mmhmm intresting! GEK 2900! Nov 16 '12 at 1:37

## 2 Answers

It is important to note that a graphical model is a very different thing than a graph of your data. What you are doing is plotting a histogram. I would drop the f# notation and just put the range of values for each bin on the x-axis, or more simply, place tick marks between the bins with the boundary values marked. Then I would put "frequency" as the title of your y-axis. Note that most software should do this for you by default.

If the homework was to create a histogram, then I would go with @gung 's approach. If the homework was to represent the data graphically, I wouldn't use bins - I would use a density plot with some sort of smoother; another possibility is a strip plot, and yet another is a box plot, but there are so few data points that a box plot really isn't needed.

• Thanks, but I have to choose between them. If I choose $10 for the y-axis, does the 10 mean the range of bins? Or it may mean 10 times the dollars appeared on x axis? Nov 12 '12 at 0:00 • I suppose per$ spent makes more sense, but both are odd. It's not "frequency per dollar spent" it's "frequency (dollars spent)" Nov 12 '12 at 1:47