I find it incredibly ironic the author who is promoting a fear based message that is weak on science is proliferating on the internet, while the article featuring the author who is questioning the science is hidden behind a paywall. As the author here states, lets' not rally around bad public science. I'm very concerned that many are going to begin using sociogenomics to dismiss issues that adolescents have been suffering from for decades, and only now via discourse on mental health awareness and acceptance, are seeking help without stigma, but will be dismissed. I am not dismissing social media's potential harm especially to the populations that Haidt suggests. The behavior we are seeing in social media is just the smoke from our cultural fire. And as Odgers indicates, it is inaccurately locating the source of the problem.
While the meta analysis she cites does say what she says about overall well being, it goes on to say that there are five categories of well being where social media had small but significant effects. That seems a bit misleading.
Considering how people seem to see the link of social media use and mental health issues with the majority of teens, I doubt that a small effect size, significant or not, matters much here. It is no at all consistent with what people seem to see. And significant p-values say nothing about the effect existing or it being or not being due to cm confounding factors or even due to p-hacking. There are a lot of small effect sizes out there with significant p-values that mean very little to absolutely nothing. And seeing how inconsistent the data is, I would not alm take it serious until they have better data that does a least a decent job at controlling for potential confounder, questionable research practices and or controlling for publication bias.
I find it incredibly ironic the author who is promoting a fear based message that is weak on science is proliferating on the internet, while the article featuring the author who is questioning the science is hidden behind a paywall. As the author here states, lets' not rally around bad public science. I'm very concerned that many are going to begin using sociogenomics to dismiss issues that adolescents have been suffering from for decades, and only now via discourse on mental health awareness and acceptance, are seeking help without stigma, but will be dismissed. I am not dismissing social media's potential harm especially to the populations that Haidt suggests. The behavior we are seeing in social media is just the smoke from our cultural fire. And as Odgers indicates, it is inaccurately locating the source of the problem.
While the meta analysis she cites does say what she says about overall well being, it goes on to say that there are five categories of well being where social media had small but significant effects. That seems a bit misleading.
Considering how people seem to see the link of social media use and mental health issues with the majority of teens, I doubt that a small effect size, significant or not, matters much here. It is no at all consistent with what people seem to see. And significant p-values say nothing about the effect existing or it being or not being due to cm confounding factors or even due to p-hacking. There are a lot of small effect sizes out there with significant p-values that mean very little to absolutely nothing. And seeing how inconsistent the data is, I would not alm take it serious until they have better data that does a least a decent job at controlling for potential confounder, questionable research practices and or controlling for publication bias.
Very important that this be more broadly known, thank you!