Post hoc ergo propter hoc. This experiment is so full of flaws it's ridiculous. Here are just some of them: * Pet food ads are extremely common. The chances o

Post hoc ergo propter hoc. This experiment is so full of flaws it's ridiculous. Here are just some of them: * Pet food ads are extremely common. The chances of not seeing them are fairly low. * The likelihood of them choosing that subject at random is also very low. It's more likely that they had seen an ad on the topic recently, but were simply unaware because of its lack of relevance to their lives. * They didn't do a placebo. They should also have discussed another common subject, away from any phones, then waited to see if they were also served advertisements about that. * Two days is arguably too long to wait to influence purchasing decisions of much needed groceries. * Confirmation bias. How many scores or even hundreds of other subjects did they see in the same period that they had not discussed. At least the "simple" in the title is not misleading.