Good shot Carlos, BUT, I think I've got a fairly simple and common-sense explanation. I scanned the article (it's huge), but it does not seem to address the obvious: Car Magazines tend to run tests, and comparisons, and reviews, when NEW models are introduced. Concurrently, manufacturers tend to launch expensive and prolific ad campaigns...when?....when NEW models are introduced. So OF COURSE there's a correlation. The launch of a new product creates both a surge of coverage, and a surge of advertising, totally independent of each other. A product of the vehicle itself, not a product of some secret collusion between manufacturers and magazines. Also, both the manufacturer and the magazine is trying to give the public what they want. That would also explain why ad content and article content share similar material. Those obvious realities can explain away Hypotheses #1 & #2 in the study, in my mind. Hypotheses #3 & #4 are fairly silly. OF COURSE a manufacturer is going to ramp up the ads for a car that wins a test or gets in the top-10. That's simple cause & effect. But that does not at all mean that the magazine mutated the results to favor a particular brand. The magazine knows that no matter WHO wins the test, or gets in the top-10, they'll likely be buying more ads in the ensuing editions. So, where's the motivation? No matter who wins, the magazine will likely get a new ad from the winner. There is no motivation to give preferential treatment to one over another, at least not enough to cause any editor to start chasing his/her tail. So while that study appears to be very thorough, I also find it to be flawed. And even still, in their conclusion, they state:
"There are limitations to the study due to its scope. First, the results cannot be generalized beyond the two magazines used in this case study. Second, correlations do not indicate causality (BINGO!!!), so they don't answer the chicken/egg question of the positive relationship. Third, the study did not attempt to measure whether the reviews were positive or negative and whether that influences advertising dollars (whoops, minor detail..
). It did, however, measure that those that won comparison features (positive coverage) were more likely to advertise."
Well, no duh.... That's an obvious conclusion, but it does nothing to indicate that it sways the magazine's judgement. I say that the magazines write it the way they see it, and the ads follow-suit. Beyond that, the correlations are purely due to new-model launches, model updates, and consumer interests that drive both vehicle and magazine sales.