Qualtrics, the world’s leading enterprise survey platform, and Evolving Strategies, a clinical data science firm, announced results today from the first clinical trial testing the effectiveness of Republican primary ads. Qualtrics and Evolving Strategies tested the effectiveness of three Republican primary ads. Some key findings include:
Days before the Iowa caucuses on February 1st, the Republican primary appears to have become a duel between Trump and Cruz. We tested three ads: one ad advocating for Trump, one ad attacking Trump, and one ad advocating for Cruz among a sample of roughly 1,200 self-identified Republican voters.
Our results suggest that Cruz’s ads alone are not enough to knock Trump out of first place with the voters he’s targeting.
This fight is a class struggle, and Cruz’s best bet to boost his numbers lies with the middle class.
Our results find that Cruz can’t win over the working class or upper class with what he’s got on the air now. In fact, blue collar voters double down on Trump when he’s attacked by Cruz.
Adam Schaeffer, Founder and Chief Science Officer of Evolving Strategies:
“Heading into Iowa, Sen. Cruz is in trouble. The standard lines of attack against Trump are effective with some sub-groups but cause other voters to double down in their support for him. Cruz would be wise to find new lines of attack while pushing out his positive ads. Otherwise, the Trump juggernaut is likely to continue.”
No. People already know what they are getting with Trump. An ad advocating for Trump does not shift votes his way. However, the same ad does move voters away from Cruz.
Not really – although Cruz has some hope with middle-class men. These men respond to Cruz's attack ad on Trump — but the attack ad does not boost support for Cruz with them; it simply causes a loss of support for Trump among middle class men. What’s more, Cruz’s attacks actually boost support for Trump among blue-collar voters — and especially among blue-collar men.
Yes, but he has to be careful with the pitch. It works with the middle-class, but it does not work with anyone else, and it actually hurts him with the upper class.
We discover the most effective messages using the scientific method — randomized-controlled trials allow us to measure a message’s impact on people in a test group relative to placebo group.
Rather than ask people what they think of a message, we observe what the message does to people when we expose them to it. It doesn't matter what people think of a message — it matters what a message does to them. We don't ask people what they think of the message we send them. Our method is like a clinical drug trial, but we test messages instead of drugs.
In a clinical drug trial, researchers don't hand a pill to a patient and ask them whether they think it will work. Instead, medical researchers measure the pill's impact on the health outcomes of a treatment group compared to a placebo’s impact on the health outcomes of a control group.
In this study, we recruited over 1,200 respondents approximating the registered voter population on age, gender, and education from an online consumer research panel. Each respondent entered the survey and answered the same questions, had the exact same experience until they were randomly assigned in the middle of the survey to view one, and only one, of four videos:
After viewing the ad, each person had been assigned, every respondent completed the same survey, answering the same questions. These questions included one where respondents were asked to eliminate one candidate, to pick their first choice for the nomination, and pick their second choice. Respondents were also asked to rate the candidates Donald Trump, Ted Cruz and Marco Rubio and to rate the chance that they would turn out to vote in their state’s caucus or primary for the Republican nomination.
We then conducted statistical analyses to determine the impact of each ad on first-place support for each primary candidate.