> “Social media research has largely assumed that [so-called] social media addiction is going to follow the same framework as drug addiction,” said Turner. Orben’s team and others argue that this is likely to be oversimplistic and are investigating whether the teenagers cluster into groups whose behaviour can be predicted by other personality traits. > It could be that, for some, their relationship is akin to a behav…
There's nothing wrong with that. But I do think it's wrong that we're not open about it, in so much that it isn't even always explicitly stated during a science education. We might say something like "all models are wrong" but I don't think that solidifies this thinking enough for novices and can create experts who still don't get it. Wherein it becomes a classic clique, where everyone knows the phrase but doesn't understand or practice the lesson.
There's some noticeable downsides too. If you're constantly not reiterating what your base assumptions are and acknowledging the limitation of those assumptions, you forget they exist. It also becomes hard to communicate the significance of your work to the public and I think in some cases is often why the public ends up calling (sometimes rightfully, sometimes not) bullshit. I think there's also a major advantage to science, in being open about these helps newcomers (such as those entering graduate school) more readily understand the limitations of the field and what work needs to be done. But I think some incentive structures we have in place don't encourage great scientific behavior in this way. Speaking of which, other metrics can even end up making bridging this multivariate gap more difficult. Such as the all too well known p-value weirdness, you may end up rejecting works making progress towards this bridging by having these arbitrary requirements rather than evaluating works with more nuance, which can even then discourage attempting to build such bridges. I guess coming full circle to: it's complicated? Like everything else lol