Repost: The Comic Potential of Two Axes

Much of scientific communication consists of throwing up a graph and then explaining it. There are some basic procedures for doing this, many of which were probably ignored by the speaker at your most recent department seminar. Don’t be that mumbledy jerkface who never explains the numbers on his or her unintelligible axes!

I suggest that you use the following graphic prompt to practice giving talks in the style of your adviser, department chair, or another charmingly be-mannerism’d colleague:

A randomly-generated graph

Reload the page to get a new one.


  1. Hob wrote:

    Correlation implies causation, right? Or is it the other way around?

    When you’re dealing with combinations like the one I just got — “Grad student office space” vs. “Unattainable hotties” — the choice of dependent variable has major consequences.

  2. Maria Brumm wrote:

    By convention, whatever’s on the Y axis is the dependent variable. However, without plotting a graph of attainable hotties for comparison, you’re not going to be able to usefully inform public policy.

  3. ScienceWoman wrote:

    Excellent – how ever did you come up with this?

  4. Lab Lemming wrote:

    Just outta curiosity, does your ad counter consider each reload to be a new page view?

    If so, you might be onto something here…

  5. Maria Brumm wrote:

    Lemming: Yes, yes it does.

    Next time I have serious case of procrastination on my hands I’ll post something similar with ternary diagrams… though I expect that may exceed the number of variables for “funny” and move into the regime of plain ol’ surrealistic garbage.

  6. Lab Lemming wrote:

    Only three?

    Make an n-dimensional one and use it for your thesis.

    btw, is it biased towards positive slopes?

  7. Hob wrote:

    Don’t overlook the earlier contributions of others in this important field. The potential is limitless.

  8. Cosma wrote:

    Further to the theme, Indexed.

  9. Ryan wrote:

    haha, I got octopodes per capita vs octopodes per capita, and somehow got:

    A) Data with no visible correlation
    B) A line of best fit which clearly does not have a slope of 1

Post a Comment

Your email is never published nor shared. Required fields are marked *