## A diagram of graphs

Jess sends along this, which isn’t a bad idea although I disagree with how it’s organized. For one thing, I think just about all graphs are comparisons; for another, I think line graphs are often the way to go, so I’m unhappy to see them in only a few of the pictures here; for another, the scatterplot-plus-regression-line, which I love, isn’t anywhere to be found. But I appreciate the thought.

### 3 Comments

1. jsalvati says:

That would make a good back page for a textbook

2. C. Zorn says:

Useful, definitely. Think I'll co-opt it for my baby-stats class next year.

That said, a few other nits picked:

– The two- and three-variable "Distribution" plots don't make sense to me (and I've never heard of a "line histogram" — it's a density plot, right?),

– No pie charts, please,

– Contour plots are generally better than their 3-D-surface counterparts,

– Matrices of plots for lots of comparisons? Bueller?

3. Nick Cox says:

I agree generally with C. Zorn.

I think to most if not all statistical people a histogram _is_ a touching bar representation of a distribution.

Evidently, a bar chart is only a bar chart if the bars are horizontal; otherwise it's a column chart. Anyone who is taught that will have to unlearn it quickly in the sense that (in my experience) most people using graphics don't ever talk about column charts, and they wouldn't define bar charts as being horizontal.

A slightly deeper point is that giving people a menu and implying that these are the choices could be unhelpful. Similarly, learning names for different kinds of charts isn't going to help graphicacy much.

Also, this looks like something from the 1950s. (Anyone remember books by Calvin Schmid?)
Quantile plots? Box plots? Dot plots?

But that's all negative. At a really elementary level this could remind students of graph forms that they had forgot about, and so be helpful.