[kerkeslager]

Do you have an alternative hypothesis for why advertising would be inversely correlated with satisfaction?

If not, it seems like we should operate on our best hypothesis with evidence for it, not speculate about unknown unknowns.

You did at one point say that advertising might not be causing the lower satisfaction, it might just be exacerbating other problems that cause lower satisfaction. While that possibility certainly warrants investigation to see what those other problems are, it doesn't have any implications for how we treat advertising: either way we should be seeking to decrease its influence on our lives.

[throwawaymath]

> If not, it seems like we should operate on our best hypothesis with evidence for it, not speculate about unknown unknowns.

Nope! We shouldn't. Speculation is precisely what we should do! That's what a critical review of research involves. We should take this as excellent motivation for further research. It could be true, but we haven't done nearly enough work to conclude as much. I'm not saying it's wrong, I'm saying it's uncertain, and it's strictly unscientific to proceed by assuming something uncertain is the truth just because it's intuitively compelling. Intuitively compelling narratives are very dangerous in science, because they make you believe you have a shortcut to the truth and they're vindicated right up until they aren't. It's not a good habit.

Happiness is incredibly complicated. Contradicting evidence and results abound across studies of life satisfaction. Measuring the impact and success of advertising itself is highly complex; the researchers have shown this correlation under their current methodology. What are we to conclude if someone uses a different methodology to study the same topic, equally as valid, and comes away with a different conclusion? That's very possible, and we can't dismiss it. The researchers themselves hedge their claims and don't come out and say they've found causation.

I don't really have a dog in the race with advertising. But I really despise this kind of reporting by HBR, because the result is threads like this one where people walk away with unproven "truths" that become part of the popular zeitgeist because they just seem to vindicate obvious beliefs. Today it's advertising, tomorrow it'll be something else. Given the replication crisis we're undergoing, we shouldn't take anything away from studies like this except that further research is required.

[kerkeslager]

> Nope! We shouldn't. Speculation is precisely what we should do! That's what a critical review of research involves. We should take this as excellent motivation for further research. It could be true, but we haven't done nearly enough work to conclude as much.

Nobody is proposing we stop researching, so let's not dignify that idea by discussing it further.

I'm proposing that as we continue researching, we proceed to take action with the best information we currently have.

> I'm not saying it's wrong, I'm saying it's uncertain, and it's strictly unscientific to proceed by assuming something uncertain is the truth...

It's not unscientific because of the nature of the truth, it's unscientific because science never tells you how to proceed, ever. It only tells you an approximation of the truth so you can make an educated decision on how you want to proceed. There is no should in science.

> ...just because it's intuitively compelling. Intuitively compelling narratives are very dangerous in science, because they make you believe you have a shortcut to the truth and they're vindicated right up until they aren't. It's not a good habit.

I'm not confused about these facts--you don't need to explain them. This is irrelevant, because this isn't why I want to take action on this study.

> Happiness is incredibly complicated. Contradicting evidence and results abound across studies of life satisfaction. Measuring the impact and success of advertising itself is highly complex; the researchers have shown this correlation under their current methodology.

> What are we to conclude if someone uses a different methodology to study the same topic, equally as valid, and comes away with a different conclusion?

If that happens, we change our minds.

What are we to conclude if someone uses a different methodology to study the same topic, equally as valid, and comes away with the same conclusion? Would you just call for more evidence?

This isn't the first study on related subjects, and so far I haven't found any that would lead me to a different conclusion.

> That's very possible, and we can't dismiss it. The researchers themselves hedge their claims and don't come out and say they've found causation.

It's also possible that the conclusions are true, and that by waiting to take action we let people suffer from the effects of advertising longer than necessary.

If we wait to take action until we're completely sure of the facts, we'll never take action, because evidence is almost never 100% conclusive.

The only information I have on this subject indicates that advertising is harmful. You are correct that this information is not 100% conclusive. However, just because we don't know something 100% conclusively, doesn't mean we should not take action on it--the vast majority of decisions in life are made with incomplete information.

We make decisions based on two things: 1. The best information we have, and 2. Our goals.

So there's basically two possibilities here: either you have some information I don't (in which case, please share that information with the class) or you have different goals (i.e. you would be pro-advertising even if it were proven that it makes people less happy).