[soreal]

See, this is the thing about population studies.

1. Millions of people live their life 2. Academics put people into buckets in order to count them 3. Other academics group that data and label those people 4. Even more other academics come along and do more groupings then write a paper. 5. A newspaper writes an article with a catchy headline and an out-of-context image with arrows drawn on it to call your attention to one correlation but not others, then proceeds to use that data to make arguments that don't actually follow

At every step along the way, a biased researcher or newspaper makes assumptions, discards outliers, and labels individuals in such a way that if we focus on their chosen pivot, we see what they want us to see.

Longitudinal studies, machine learning, and more are all improvements on the current shaky process. I just hope they continue to catch on despite being tougher to do.

[peterkshultz]

I agree with your take on newspapers: they try and capture the attention of a layman audience. As such, their work can be simplified to the point of impropriety from time to time.

I'm less inclined to agree with your opinion on academic research, however. Systems of peer review--and softer systems, like those of repute amongst colleagues--are in place for the express purpose of eliminating biases and errors across research works.

In the social sciences, there is a great amount of care that goes into ensuring that: 1. Data is handled properly 2. Future inquiries are sound

Whether or not that care happened in this body of work is up for debate.