Statistics - how they can so easily mislead

On Thursday September 8th 2011, the Guardian newspaper published a blog that included a table listing the % of votes at Westminster that MPs and members of the House of Lords had missed. For Fabian it showed a figure of some 73%.

This is a particularly stark example of the way that statistics can so easily mislead. The statistics were specific in that they covered a defined period when the two Houses in Parliament were sitting - January and February 2001. This was during the time that Fabian was seriously ill with kidney stones and spent time in hospital for two operations.

In fairness to the Guardian, the author of the original report on MPs attendance, Polly Curtis has responded apologetically to Fabian when the sitiation was pointed out to her.

Dear Mr Hamilton,

I wanted to write to personally apologise for the way your name appeared in the blog I wrote on Thursday about MPs' vote attendance rates. I was aware of the potential for such problems and had spoken with party sources to try and work out whether there were any individuals on the list who had circumstances we should know about and we did treat the entries sensitively from the outset for some of your colleagues in similar situations. Unfortunately no one mentioned your name, though prior to you writing in a reader of the blog had emailed me suggesting there might be a problem and I had asked for your name to be removed. For a technical reason that I cannot fathom there was a delay in that happening until last night. Your entry has now been removed and, after discussion with our readers' editor who deals with complains against our journalism, I have added a note to the blog to apologise to you for its original inclusion. It means that anyone coming to back to the blog will read the correction where the mistake was originally made. I write about this potential problem at several points in the blog and in response to comments from readers and I do hope you understand that it was not our aim to be unfair in the application of this data.

Kind regards,

Polly

Fabian's protest was not done just for himself as he pointed out to Polly that some other colleagues at Westminster had also been ill during that period and that there are many valid reasons why MPs have different voting records. Where issues are not contentious what is important is that a vote is taken as part of the process of the law being changed and so it would be a poor use of MPs time if they were all required to turn up every time.

Fabian said later that in this case it was not a case of 'Lies, damn lies and statitics' to quote the famous author Mark Twain but rather a reflection of the kind of situation parodied by the Scottish poet Andrew Lang, 'An unsophisticated forecaster uses statistics as a drunken man uses lamp-posts - for support rather than for illumination.'

Fabian has no objections to any figures on MPs work such as attendance and expenses beng published; such information rightly belongs in the public domain. However, those who publish should do everything possible to ensure that information informs but not misleads. The subtitle of the Guardian blog read, 'Can we find Britain's laziest MP or Lord? See what the data says?' But the data alone did not answer that question and this kind of reporting can be seriously misleading. This was certainly the case for Fabian in respect of his incapacity last January and February.

We are very pleased to note that the Guardian was concerned to correct the public record as soon as the matter was pointed out to the editors. To see the statement - Click here