PolliGraph Debate Drilldowns. Part 1: The Overview

To the joy of some and the chagrin of others, the stars of last week’s historic health debate between Rudd and Abbott were definitely the competing worms.

As good fortune would have it, Roy Morgan — the owner of the Morgan Reactor technology that powered Channel Seven’s PolliGraph — has shared respondent level data with Pollytics and Crikey to let us bring you the most arcane secrets of what was the most widely witnessed political focus group of the last few years. So a big thanks goes out to them for letting us see the results of what is some of the most spiffy focus group and audience response technology on the planet.

Over the coming days, we’ll bring you all sorts of demographic breakdown goodness of the debate, focusing on what messages pressed the buttons of which cohorts, how different angles pursued by Abbott and Rudd played out with different cross-tabs, and we’ll also take a look at a few results that may surprise you.

Today, however, we’ll take a bit of an overview of the technology and a quick squiz at some basic demographic patterns.

As a direct result of the very short period of time that was available to organise an audience panel to participate in the PolliGraph, the partisan balance of the audience had a slight tilt towards Labor, giving the responses we witnessed on Channel Seven during the debate a small but significant lean towards the ALP. If the debates were announced a few days earlier, there would have been enough of a lead time to get a properly balanced panel – often, time is the killer with such things.

In numerical terms, Labor vs. Coalition voters split a tad over 60/40 towards the ALP. So while the broad, average level of the audience response might lean toward Labor slightly, the changes in that response over time are still politically neutral. We can also remove that partisan lean entirely – which you’ll see in action later – by looking at Labor and Coalition voters as separate, individual cohorts.

Even though it was fairly obvious that Rudd won the debate, as a result of this slight lean, the PolliGraph data suggests that the PM didn’t clean the Opposition leader’s clock to the extent that the various worms initially conveyed. Rudd still won conclusively, but it wasn’t a complete train wreck for Tony Abbott.

To give an idea of the sorts of things we’ll be doing and the cohorts involved, if we have a quick squiz at the audience responses of the opening statements of the two leaders, and run the PolliGraph by both voting intention and gender, we can see how the basic patterns and leans play out for each leader in each demographic. These are just still images – in later posts we’ll run video as well.

First up is Rudd, by gender and voting intention. Each image represents a two minute block of the opening statement – so this is what the first four minutes looks like: (just click to expand)

It took the audience a little while to react to Rudd, but when they did a pattern emerged that became a regular feature of the entire debate – Rudd’s words and messages were most approved of by Labor voting women, least approved of by Liberal voting women and the men fell somewhere in the middle.