Wayward Woodward

Arianna Huffington has written a scathing commentary on Bob Woodward's gentle treatment of the Bush administration and its Iraq war propaganda.

To bar or not to bar?

Bar charts are sometimes used to show percentages, like this:
Looks like a trend, right? The top of each bar gives a point estimate of the percentage at each time. But to evaluate whether the apparent trend is more than the play of chance, we need to consider the precision of the estimates. So I often recommend showing confidence intervals around the point estimates, like this:
The overlap of the confidence intervals is considerable, which suggests that the data are consistent with there being no trend at all. In SPSS, the chi-square trend test ("Linear-by-Linear Association") gives a 2-sided p-value of 0.229, so it's clearly not statistically significant. (The reason is that the proportions are based on very small denominators.)

At a blog called Junkcharts there was a recent posting titled When not to use bars. I added a comment (which I've edited a bit below):

"I think that one of the reasons the bar chart is so popular is that it paints broad strokes of ink (particularly striking when color is used), giving the figure a kind of visual punch. A bar chart can be seen from halfway across a room, whereas the traditional figure favoured by statisticians nearly disappears (admittedly I'm not wearing my glasses, but I think the point holds). However, one could achieve a similar visual effect to the bar chart using vertical colored boxes with light horizontal lines indicating the point estimates."

Here's what I meant:

Of course it's a rather unfamiliar display. I'd be interested to hear other people's thoughts on this.

The vocal stylings of Jean Chrétien

The beginning of this month witnessed the release of the Gomery Commission's Phase 1 report. Ever the trickster, Jean Chrétien's response is to attempt to charm us with music.

The waiting game

Cataract Surgeries in Ontario 2003/2004

"For the first time in this province's history, Ontarians can now find out how long they can expect to wait for key health services at hospitals. This site contains the most up-to-date information about wait times in Ontario, with a breakdown for five specific health services by individual hospitals that provide those services."

For those of us who live in Ontario, it's especially interesting to look at the "Wait Times in Your Area" -- you just type in your postal code and see the median wait time to get an MRI at nearby hospitals (as well as the mean and the 90th percentile). There are a few other goodies as well.

One section that caught my eye was "Myths About Wait Times". Whenever I see the word "myths" used like this (particularly in the dichotomy "myths and facts about X"), I have a kind of intellectually aggressive reaction, and I immediately look for flaws. They're usually very easy to find and this case is no exception, but I'll leave that to the gentle reader (whom I encourage to leave a comment).

Word of the day: floccinaucinihilipilification

From a recent commentary in the Journal of Clinical Epidemiology (I put some links in):

"... Petticrew and Egan comment that: ‘‘There is a widespread assumption that what systematic reviews are best at is assessing ‘the evidence’ and concluding that it is too little, and anyway what there is of it, isn’t very good’’[1]. They highlight a useful 19th century word for this activity - floccinaucinihilipilification - which, according to the Oxford English Dictionary means ‘‘The action or habit of estimating something as worthless.’’ "

[1] Petticrew M, Egan M. Relevance, rigour and systematic reviews. In Popay J, editor. Moving beyond effectiveness in evidence synthesis: methodological issues in the synthesis of diverse sources of evidence. London: NICE, in press.

Paint with all the colours of the wind

Here's a function I wrote in R to help me pick colours when producing beautiful graphs:

showcol <- function(col) {
co <- colors()[grep(col,colors())]
thickness <- floor(400/length(co))
par(mar=c(1,8,1,1))
par(las=1)
plot(0,0,type="n",xlim=c(0,1),
ylim=c(1,length(co)),axes=FALSE,
xlab="",ylab="")
mtext(co,line=1,side=2,at=1:length(co),cex=0.7)
for (i in 1:length(co)) {
segments(0,i,1,i,col=co[i],lwd=thickness)
}
}

To produce the figure at right, you simply call

showcol("brown")

Some of those don't look all that brown to me ...

P.S. I couldn't figure out how to get Blogger to keep my indentation in the code above, but at least you can copy and paste this directly into R and it works. And I got to learn all about CSS.

The incredible lightness of being

To be or not to be?

Gotta love PowerPoint

I'm preparing a talk on graphs and tables in medical journals and I happened on something by Edward Tufte:

"Applying the widely-used default designs for statistical graphics in PowerPoint [to a table of long-term survival rates of cancer patients] yields these analytical disasters below."

Tufte's assessment:

"Everything is wrong with these smarmy, nearly unreadable graphs: incoherent, uncomparative, low data-density, encoded legends, color without content, logotype branding, chartjunk, indifference to content and evidence. Chartjunk is a clear sign of statistical stupidity ... Poking a finger into the eye of thought, these graphics would turn into a particularly nasty prank if ever used for a serious purpose, such as cancer patients seeking to assess their survival chances. To deal with a product that clutters and corrupts data with such systematic intensity must require an enormous insulation from statistical reasoning by Microsoft PP executives and programmers, PP textbook writers, and presenters of such chartjunk."

Is that clear enough?

A friend was telling me about a philosophy course she was taking: "When I got to Kant, I couldn't."

Pythagoras meets flash

I've been learning how to program Macromedia Flash. Here's one of my first projects, based on a multiple pictorial proof by Paul Vjecsner:

Getting started

The photo on the left seems like a nice way to start. It was created using a zany method that produces some beautiful results.



I'm sure someone's written a really good piece on why blogs have become so popular (if you know of one, please put a link to it in a comment). It seems to me there are some interesting historical points (in terms of the printing press, Marshall McLuhan, etc.) as well as sociological, political, even spiritual aspects.

My reasons for having a blog are as follows: 1) It's a very convenient way for me to keep an organized log of interesting stuff that will build up over time into a kind of database; 2) Someone else might find my stuff useful, informative, or entertaining; and of course 3) It's fun and creative.

Now, here's an interesting blog.