Anscombe’s quartet and the importance of graphing your data

I’ve stalled a little bit in my thesis writing – I’ve only got reanalysis and in-depth rewriting to be done on this chapter, which I’m not keen to do at this time of night. So I thought I’d bash out a quick post before the new year.

When dealing with sets of data from familiar experiments, it might be tempting to throw the numbers into your favourite statistics software package, and report the coefficient and p value. But researcher beware! Strange things may be hiding in your data… Anscombe’s quartet is a fine example of this. The quartet is four sets of data that have the same sample statistics (mean, variance, correlation coefficient and regression equation), but when graphed, they are clearly very different.

The distributions in four very different datasets produce the same statistics

Anscombe's quartet plotted, from Wikipedia

The quartet is only an illustrative example of what is possible; the Wikipedia article has links to other similar data.

But graphing your data doesn’t just guard against mistakes, it can also allow you to see patterns in your data that you hadn’t thought to look for. If you use R, there are plenty of snippets of code that make a summary plot of data, with frequency distributions and Q-Q plots. So give it a go. Work with your data from the bottom up – you never know what you might find.

P.S. Happy New Year everyone – There’s a surprise coming for Neuromancy next year…

Is science turning authoritarian?

Going through the things I’d bookmarked to write about, I found an article in The American entitled “Science turns authoritarian”, in which it is argued that science has let itself become co-opted by politics, has begun to espouse view points rather than describe the world, and that as a result science is losing its credibility.To support this view, the following graph of frequency of so-called “authoritarian” phrases is presented.

The incidence of phrases such as "science tells us we should..." has increased since 1990.

 

I think two points are worth making about this, one of methodology and one of interpretation. Firstly, the graph is based on the frequency of these phrases in science reporting, rather than from scientists themselves. Whilst scientists could also be expressing similar attitudes, the position that these data support is that science is perceived to be to more authoritarian because those who report on it are perceiving it to be more authoritarian. Also, although the article suggests that if the increase were just a result of an increase in print volume then the frequency of all the words would increase equally. I’d argue that even if there was an increase in authoritarian tone, some of the phrases are going to be more commonly used than others. I’d also venture that use of some of the phrases (especially “science requires”) aren’t necessarily authoritarian – “a successful career in science requires dedication”?

Secondly, I think that the implication in phrases such as “science says we must” is a conditional statement that we must do X if we want Y. Where Y is generally something like the continuation of life as we know it. The article even cites some examples e.g. “For example, science tells us we must reduce our global greenhouse gas emissions to prevent dangerous climate change.” You could read that as saying emissions must be reduced if we are to prevent climate change. Even if you read it as making the assumption that we should want to prevent dangerous climate change, it doesn’t seem like an unreasonable assumption.

Which leads on to a third point. I don’t think scientists are capable of, or should be entirely objective in the sense of robotic detachment from the implications of their work. If reasearch strongly suggests that if we don’t drastically change the way we are living things will get terminally hard for a lot of people, you’d have to be inhuman not to feel like it’s something that we should probably be doing. And it’s probably something we should be doing more of. I think it’s particularly telling when the article mentions that people don’t take kindly to being told what to do “particularly when they’re unable to evaluate the information that supposedly requires them to give up their SUV, their celebratory cigar, or their chicken nuggets”. In circumstances where the information clearly requires a course of action, but only those immersed in the field have the ability to evaluate the information, it seems almost reckless to not give any context.

See The Authoritarianism Claim at Only In It For The Gold for a similar but more concise take on things.

Follow

Get every new post delivered to your Inbox.