BBC News
Launch consoleBBC NEWS CHANNEL
Page last updated at 10:53 GMT, Friday, 28 August 2009 11:53 UK

Understanding rape statistics

Woman with her head in her hand
More or Less
Friday, 28 August 2009
BBC Radio 4, 1330 BST

The conviction rate for rape in England and Wales is notoriously low.

But how much does the number - 6.5 per cent - really tell you?

Ruth Alexander explains where the figure comes from and asks whether England and Wales really are the worst in Europe for getting justice for victims.

We give the full picture - and explore why cross-country comparisons cannot be taken at face value.

Degrees of Wealth

Students from Brimingham University at their graduation ceremony
If this lot studied maths, they could expect fat pay cheques

The government says that graduates tend to earn substantially more than people who have A-levels but did not go on to do a degree.

In fact, it says, projected over a working lifetime, the difference is something like £100 000.

But the claim masks a great deal of variation.

The government does not always mention that the extra money a graduate can expect to earn depends on the subject they study, the university or college where they do it and in which region they end up working.

And some degrees might leave you likely to earn less over the course of your career than you would have done had you not gone to university at all.

Trouble Mounting with Double Counting

A black and white photo of a man with a stack of books
Statistical meta-analysis combines results from several related studies

A meta-analysis takes the best scientific studies addressing a specific question and combines their results to produce the most reliable answer.

It represents the strongest kind of scientific evidence.

But we hear from a professor of statistics who thinks the way they are compiled and presented can lead to basic mathematical mistakes.

The Beauty of Data

Section of image showing 24 hours of flight paths over the north east of the USA
There's a new trend for using data to create interesting images

The increasing availability of data over the internet has fuelled new experiments in the field of "data art".

Presenter Tim Harford explores some recent innovations.

And he finds out whether anyone can make art from raw numbers.

Other links


BBC Radio 4's More or Less is broadcast on Friday, 28 August at 1330 BST and repeated on Sunday, 30 August at 2000 BST.

Subscribe to the More or Less podcast.


You can e-mail us by using the form below:

Name:
Your E-mail address:
Country:
Comments:

Disclaimer: The BBC may edit your comments and cannot guarantee that all emails will be published.

Your Comments

Understanding Rape Statistics

My view is that rather than encourage the reporting of rape, the unique way the published rape stats are calculated would discourage reporting of attacks. If victims feel that after the attack and the necessary physically and emotionally invasive investigations, they have a less than 5% chance of finding justice they may well decide not to inform the Police whereas a 46% chance (if calculated consistently with all other crime figures) may well encourage reporting. Ian Ford

Why, when statistics regarding rape convictions are discussed, is it assumed that those figures indicate that a high proportion of rapists are escaping convictions, when surely, the high proportion of allegations of rape which result in no conviction might be interpreted differently - that in fact, there may be a disproportionate number of false allegations? Steve

Graduate salaries

By definition, lifetime earnings can only be assessed at the end of a life. Therefore the people whose earnings are being used for assessments are those who took their degrees long ago, in an entirely different employment world. Nowadays, when many graduates may face years or lifetimes either unemployed or stacking shelves, it must be preposterous to assume the same earnings premium enjoyed by their grandparents' generation. Michael Oyle, Ireland.

Your statistics seem to take no account of the graduate debt, likely to be £20,000 - £30,000. This means that the non-graduate has an opportunity for a settled career and lifestyle while the graduate is still repaying this debt. Terence Neal

In the 21st Century there are very few young people with good A-levels who do not go on to higher education. By definition they are untypical and therefore do not provide a good basis to estimate what graduates would have earned had they not gone to university.

With respect to the variation by subject the comparisons are made with the same A-level qualifiers who do not go to university. As if the A-level qualifier who would have done maths had he or she gone to university were the same as the A-level qualifier who would have done Art.

The comparisons of the premium between universities do not require comparisons with A-level qualifiers, and there are studies that try to compare them. The problem here is that for contrasting institutions the numbers of similar students are very small. They are small enough to be data errors. John Thompson

The item on graduate salary premium did not cover one relevant point. Presumably the figures quoted extrapolate from a number of past years when the proportion of graduates was far smaller. As we get close to 50% of the cohort, what is the premium likely to be? Gordon Graham

Meta analysis

I have worked as a scientist and have peer reviewed on a number of occasions. To do it properly is a very onerous task and from my experience, the standard of submissions was mostly poor and on occasion possibly fraudulent. The pressure to publish, the explosion of journals and the cut-throat nature of academia results in a number of questionable practices. Having worked in science, I treat most published work with suspicion until proved otherwise. Name withheld

Data art

In Britain we do not need 'data artists' to visualise data in order to understand it or use it in creating patterns. This is because we have geographers. Geography is not a taught subject in USA schools so maybe they have lost out on spatial perception. Both the air traffic over the States and Napoleon's journey to and from Moscow are varieties of flow line mapping that can be based on satellite images of the earth or more conventional base maps. Perhaps you can devote part of a programme to the increasingly important statistical mapping element of geography. Angela Bailey

How can you possibly have covered data/art without mentioning Processing? It's what Aaron Koblin used to make the images that he created of the flight patterns. Chris Lynas

I have not seen this realised but the first instance of the concept that I know of was in Douglas Adams's "Dirk Gentley's Holistic Detective Agency" written about 15-years ago. Are the mathematicians still having to play catch-up with the artist philosophers? Richard Mullen

Ashes statistics

I was fascinated by your piece on the Ashes yesterday. I worked as a statistician for many years and I had got so irritated by the claim that somehow Australia had statistically "won" the Ashes that quite independently I had removed the results for the first test and shown, like your piece that the performances were broadly equal after that, with Broad featuring near the top for both bowling and batting - the commentators had wanted him dropped!Michael Stock



SEE ALSO
Data project awarded 10m funding
26 Aug 09 |  Cambridgeshire
Graduates 'reap social benefit'
18 Aug 09 |  Education
New students 'face 23,000 debt'
18 Aug 09 |  Education

RELATED BBC LINKS

RELATED INTERNET LINKS
The BBC is not responsible for the content of external internet sites


FEATURES, VIEWS, ANALYSIS
Has China's housing bubble burst?
How the world's oldest clove tree defied an empire
Why Royal Ballet principal Sergei Polunin quit

banner watch listen bbc sport Americas Africa Europe Middle East South Asia Asia Pacific