大象传媒

Cambridge Analytica became involved in a scandal in the spring of 2018 when it was revealed that it had collected the personal information of millions of Facebook users without their knowledge. The firm claimed that it could use this information to influence elections by "micro-targeting" voters - giving them messages tailored to their individual personalities. There were claims that it had helped swing the US presidential election for Donald Trump, and allegations that it had been involved in the Brexit campaign. If true, it seemed that Cambridge Analytica had discovered a way to mess with voters' heads by identifying and then exploiting their secret hopes and fears. It seemed sneaky, if not downright sinister.

But leaving aside the specifics of what Cambridge Analytica actually did, how well does the science behind the alleged method stack up? Can the population be sliced, diced, and targeted through sophisticated "psychographic" techniques. in SLICE, Jolyon Jenkins investigates by breaking it down into five areas: Selfhood, Likes, Inclinations, Convincing, and Elections:

Selfhood: psychologists claim that someone's "personality" can be measured according to five independent factors: Openness, Conscientiousness, Extroversion, Agreeableness and Neuroticism. These are known as the "Big Five". Cambridge Analytica relied on the Big Five for its method. But can we all be boiled down like this? How did we end up with five factors? The bizarre history of personality testing is filled with larger-than-life egos, dubious hypotheses and - some say - questionable methodology. Are the Big Five dominant today just because five is a handy number, and because it lets all psychologists use the same scale?

Likes: When two Cambridge University psychology researchers started using Facebook to do personality tests, little did they know what they would unleash. Not only could Facebook be used to administer personality questionnaires, but they discovered that people's Facebook behaviour, including their "likes" could be used to predict personality. And, they claimed, computer-based personality-based judgements are more accurate than those made by humans. Cambridge Analytica took the idea and ran with it. But now, even Cambridge Analytica's own data expert says that the claim is overblown.

Inclinations: Does someone's personality tell you anything about how they are likely to vote? People assessed as "Open to new experiences" tend to have more liberal political opinions, but isn't that exactly what you would expect? During the Brexit campaign, areas of the country where the population scored low on "openness" were significantly more likely to vote "leave". But why are those areas more "closed" in the first place? Could there be a genetic factor, or is it the environment? Could it even be that areas that have been hit by infectious disease in past centuries have a more "closed" population because avoiding strangers was the best way to avoid infection?

Campaigns: If your Facebook behaviour reveals whether you're an extrovert or introvert, neurotic or stable, agreeable or unpleasant, can these results be used to get you to change your behaviour? Researchers at Cambridge found that you could sell more cosmetic products to extroverts and introverts if you gave them messages targeted to the particular personality. Cambridge Analytica claimed that in America they could get a pro-gun rights message through most effectively to neurotic people by targeting them with a fear-based message, whereas conscientious people would be better influenced by a message that focused on tradition and stability. True or false?

Elections: Whatever Cambridge Analytica did or did not do, the data that Facebook and other big data companies have can almost certainly be used effectively for political campaigning. We speak to people involved in the last UK general election about how the data harvested by Facebook itself - not Cambridge Analytica - was used to deliver targeted messages to particular groups of voters. How Labour sent messages to pro-Brexit Labour supporters to reassure them that Jeremy Corbyn was not a closet remainer. How the two main parties bid against each other for such Google search terms as "Dementia Tax" in an attempt to reach wavering voters. Targeting voters through their digital footprint seems here to stay. But is that necessarily a bad thing? Some of the original Cambridge university researchers argue that, in an era when people are disengaged from politics and ill-informed, microtargeting voters to connect with their particular concerns could be a useful way to get the population re-engaged with the political process.

Presenter/Producer: Jolyon Jenkins.