Not really, it’s just a catchy headline to point out the problem with correlation being confused with causation. Here’s a plot I just made with the current sporting event (better not name it in case I get sued) in mind:
There’s a strong correlation between smog and human rights abuses, but we can’t say that one is causing the other!
Here’s a quick “how to” because I don’t have another blog to put it on.
If you are making a poster with web browser screenshots on it, they will look poor quality when eventually printed out at A0 size. This is because unless your screen is absolutely huge, there aren’t enough dots there to look good on a poster. Ideally you should work with vector graphics formats (like EPS, PDF) in a proper drawing program like Adobe Illustrator (since PDF is imported as an ugly low-res bitmap into Powerpoint), but if you can’t do that, then the following may work for you.
Read the rest of this entry »
For some reason, I thought it would be interesting to follow “hay fever” through the days of the week. I couldn’t think of another good query to compare it against so I typed the same thing without quotes. The result was quite interesting:
Forget the US elections, things are hotting up in London’s mayoral race. Any PR is good PR for Boris Johnson it seems. His web visibility is rising faster than Ken Livingstone’s…
Well you heard it here first. April Fools and gooseberry fool are both on the up.
April Fools is getting more popular pretty damn fast actually (try some other controls like silly fool).
Reading The tyranny of pink in the Guardian today I had to try this one out:
The thrust of the article, that cutesy pink stuff is increasingly turning our daughters into zombie consumers, is true. Although pink was all over Compare Stuff News first…
Another time trend plot, showing Barack Obama’s relatively faster rise to fame compared to the Clinton brand. Perhaps no surprise as I (in the UK) had never heard of him until the current race.
Let’s see if talk of house repossessions is on the rise, and let’s use a similar phrase as a control:
It seems to be true, although no meteoric trends here.
The plot below shows the biological sub-disciplines leading in the area of open access publishing (click on the plot for readable x-axis labels):
Bioinformatics has by far the largest amount of “open access” chatter. In purple you can see the relatively flat distribution of “journal” as a control. As a bioinformaticist myself this seems to make sense, as the leading figures in the open access movement come from this field. However, other communities (physics, computer science) have been operating an informal open access model for a lot longer (preprint servers and/or widespread preprint availability).
A lot of new x-axis categories have crept in over recent weeks – I’ll try to feature some on the blog. How about the old homepage faithful stressed/relaxed with respect to major life events…
That suggests that being pregnant and starting school are the most stressful events, but wait, let’s try it with “stressful” too, just as a sanity check:
Read the rest of this entry »