A statistical plot demonstrating quantity and std (or mean)

What kind of statistical plot do you think can demonstrate both the quantity and std (or mean)?

For example, I have a dataset of a large population in different ages. I used clustering method to separate the population into ten groups based on the Euclidian metric of age. Now I want to examine the population quantity (counts) and age std (or mean) in each clustered group. Ideally, the std for each group will be very small, but their mean increases monotonically.

Answer 1

[Image Analyst]

What about a box plot http://www.mathworks.com/help/stats/boxplot.html?s_tid=srchtitle

[Kyle Wang]

Thank you for the answer. However, I think boxplot doesn't demonstrate the quantity (counts or percentage) of a population? It shows the distribution.

[Image Analyst]

A histogram displayed as a bar chart shows the counts or percentage. It would be like the sample I showed with bars going up to every red line. However a bar chart doesn't show the distribution of each category. If you use cumsum(), you can get the percentage (cumulative distribution function) going across categories or bins - not sure if you wanted the standard deviation across bins/categories or within a bin/category.

Anyway, I'm really confused why you say the box plot doesn't show the quantity when the level of the red line is exactly the quantity (or counts, or percentage if you normalize it). Exactly what do you think the red line represents, if not the quantity?

[Kyle Wang]

Thank you for the explanation, Image Analyst. The red line in the boxplot I think is median, not counts (the number of samples in the group).

[Image Analyst]

You're right - I hadn't read the help completely. I guess the mean would be at the halfway point in the blue box. I guess you could draw a line for that if you wanted - unfortunately I don't see an option to have the mean line in there explicitly.