I stride into the classroom, ready for a challenging session with 4C on Statistics, but I think I’ve got a way of presenting the topic in a way that will hold their interest with the forthcoming election.

I cough, wait a few seconds for a hubbub to die down, and then commence. “My once favourite newspaper, the Telegraph, now my most despised rag for having sacrificed its’ true conservative tendencies to back David Cameron and his ragbag collection of fellow…” I pull myself up, realising I am over-stepping the mark on the political bias guidelines.

The class titters. I cough again and resume. “The Telegraph has gone in for Lies, Damned Lies, Statistics… and Misleading Graphs… big time.”

I’d like to point out a spectacular example, where a seemingly innocent and “balanced” graph in fact speaks like man with forked tongue. Yesterday, there was an article The Green Party is UKIP for young people which sported this… I fire up the overhead projector to display a graph:

UKIP Greens Young Old

“So far, so good,” I continue having given them a chance to absorb it. Jones Junior shouts out, “My Dad’s voting UKIP, but mum’s voting Green.”

I shake my head sadly as his interjection gets a hubbub going and bash on the blackboard with my pointing stick. “Has the graph taken you in? Do you think that electoral support for UKIP amongst older people is balanced by support for the Greens by younger people?”

Some of them nod their heads, so I shake my head sadly and whack on the blackboard with my stick. “Right, you lot, look at this.”

For a start, let’s take a look at the population estimates from the Office of National Statistics. To save you all the hard work, I’ve downloaded a file and aggregated the results for each of age groups in the Telegraph chart. Here’s the result:

Age Group Population
18-24 5.20m
25-39 11.27m
40-59 15.27m
60+ 13.03m

“Are you beginning to see what I mean?” I look around the class, there’s a glimmer of dawning recognition on the faces of those who can do mental arithmetic well, without their computers.

The 18-39 total is around 16.5 million people, but there are 28.3 million people over 40, nearly twice as many.  Then take a closer look at the graph – 18% of the larger older group (based on the rather UKIP-unfriendly YouGov poll they are using) would vote UKIP, whereas around 13% of the smaller younger group would vote Green.

I tap on the left-hand blackboard as I move over for emphasis. “It doesn’t end there,” I emphasize, “take a look at the “likelihood to vote” figures. It’s actually quite difficult to nail these, but here are the actual figures from the 2010 General Election:”

Age Percentage Voting
18-24 51.8%
25-34 57.3%
35-44 64.4%
45-54 67.5%
55-64 69.8%
65+ 74.7%

It’s not a direct alignment with the YouGov age groups, but we can interpolate to come up with some reasonably accurate figures.

I now move to the third blackboard, which is nice and clean, and chalk up some headings and the left hand row, a template for a complete table:

Age Group People Voter % % Green % UKIP Votes Green Votes UKIP
18-24  ? ?
25-39  ?  ?
40-59  ?  ?
60+  ?  ?

“Right,” I announce, “I want 6 volunteers, one to complete each column of the table from the data I’ve already given you.”

They’re a bit coy to begin with, but slowly they come up, the smart ones taking the first 4 easy columns, leaving the last two more difficult columns. Eventually, we get there with the help of calculators on their muted mobile phones:

Age Group People Voter % % Green % UKIP Votes Green Votes UKIP
18-24 5,200,000 52.0% 19% 3% 513,760 81,120
25-39 11,270,000 60.0% 10% 10% 676,200 676,200
40-59 15,270,000 67.5% 6% 16% 618,435 1,649,160
60+ 13,030,000 73.5% 4% 20% 383,082 1,915,410
TOTALS: 2,191,477 4,321,890

There you have it. In fact, we are back to the “face value” Opinion Poll result of 8% Greens and 16% UKIP. However, what this proves is that it is like the art of the magician – distraction. The reader’s eye can be drawn away from the factual truth (UKIP have twice as much support in the poll as Greens) with a snazzy headline (The Green Party is UKIP for young people) and a misleading graph.

I could go on and draw your attention to the application of this topic to the polls themselves, YouGov being one of the worst when it comes to offering misleading political poll results, but that will be the subject of another lesson.

“Class dismissed,” I announce, as they stampede for the door and the mid-morning break.

Print Friendly, PDF & Email