Vaguely speaking, my profession has a lot to do with understanding the "why" of politics and requires a lot of statistical knowledge. Long post incoming.
There was a time when I'd try to educate people to inform their misconceptions but I have to restrain myself from doing it often here. You can't teach people who don't want to accept new information, and the culture of self-righteousness is pretty deep-seeded in this community, which I ashamedly assume at times too. Thing is, when you actually know a thing or two, your expertise is treated like any other opinion (then again, I consciously avoid advertising my credentials).
RE: Politics:
Pretty much every politics thread starts with shitposts which set the tone for the rest of the discussion. Unmoderated, editorialized thread titles about political news encourage this shitposting. The result is threads full of what some call "hot cognitions" (many different terms for this); Essentially, highly emotionally charged responses that lack critical insight. Making things worse is that your average poster in these threads is pretty uninformed, which is understandable given that most people in an electorate are, except for the fact that their opinions are often passed on as knowledge and recirculated throughout the community.
The result? Political discussion here is basically a liberal echo chamber that *actively and structurally discourages* critical thought. There probably isn't better evidence of this than the pervasive dehumanization of conservatives. I'm politically in-line with most folk here but the open promotion and differentially-moderated hatred towards conservatives is a really, really, really, really, really bad thing. It's antithetical to the goals of progressivism, promotes polarization, discourages empathy for people with opposing views (who, according to research, actually do have similar policy attitudes than they think and would be better served as allies) and has played out really, really, really, really, really badly when you look at the consequences of power shifts in history when competing groups have been demonized. These are things people should know are bad and would absolutely know how bad they are if they were informed. Instead, people basically use conservatism and fascism synonymously here.
Based on my credentials, I imagine I'm the only or among a select few people on this forum that might be able to answer the titular question in the "What's the appeal of conservatism?" thread, but I had to close the tab to stop myself from doing so. It's not worth it, there is no good faith argumentation that can be had there. A democracy is dying before people's eyes and they're musing about whether voting for Joe Biden is the right thing to do. People are willing to die on their hill even if it means they let a global superpower succumb to fascism.
Sometimes I snap and go in on how absolutely stupid some of these perspectives are but.. my knowledge also informs me of how and why people came to these conclusions. Why they might dehumanize, why they might not see the forest through the trees, why they might not have the psychological/emotional bandwidth to be critical, why people who have been historically targeted and persecuted by conservatives would be contemptuous, why it's okay/normal for them to be uninformed and that's not really their fault, we're living etc.
It's incredibly frustrating because there's no effective method of communication knowledge here, people are seemingly and unknowingly adopting attitudes and behaviours that are counterintuitive to their beliefs in the long run (which, I might add, conservatives are frequently criticized for doing), and it feels like moderation is effectively preserving the echo chamber that makes people less open to communication. I'm hoping that I get some free time in the future to make a more well-thought out post in the constructive criticism thread that mods have opened up (which gives me hope and I love that mods are doing because this doesn't mean that they aren't well-intentioned).
RE: Statistics:
Statistics are really hard to interpret and effective communication of science is incredibly difficult, bound by the limitations of the data and statistical tests used. Interpretation is a big part of research. Picture this: Your statistical test usually pumps out numbers that only a researcher with multiple post-secondary degrees and years of specialization is capable of and responsible for accurately understanding. Then, if it gets publicized, the media has to take that understanding and effectively convey it at a ~6th grade reading level in a 280 character tweet for it to have any chance of someone outside of the academe reading it.
At every step of this communication process is error. In fact, this game of broken telephone extends to the statistical tests AND the design of the studies you are doing the statistical tests on. Error is built-in to every step. This is why it is so hard for your average person to read a research paper. Accordingly, this is also why it is so hard for your average person to criticize a study they see in the news. The people who can do this aren't average people, they're peer reviewers.
So what do people do wrong here, exactly? First, people treat data as truth. This is wrong, because, as I just mentioned, data always has error, and always must be interpreted. After the 2016 U.S. Presidential Election, people became disillusioned with polls and statistics because they treated the data like it was a crystal ball that would tell them the future to calm their anxiety and give them a sense of security. Again, data is not truth. It is an abstraction of and/or incomplete truth, and this will forever be the case until we can literally predict the future. The uncomfortable truth of science is that it is fundamentally imprecise and constantly requires ripping up what you thought was right and replacing it with something tangibly better but still imprecise. This is especially the case for social sciences. Repeat after me, data is not truth.
The other thing that people use certain "go-to" criticisms to disparage studies that don't line up with their belief systems. A big one here is "small sample size." There are so many problems with this, I'll just go into a few. First, sample sizes are determined by so many factors, most often than not how much money they have to run that sample. Second, sample sizes are really only too low if they don't sufficiently "power" the statistical tests that are run on them. To determine this, researchers sometimes run a whole separate set of analyses just to determine how many participants they need, called a power analysis. Third, in many cases you should probably be more worried about the representativeness of the sample than the sample size. It may have thousands of participants, but are they all white college students? How applicable are your findings if the data represents such a specific group of people? Fourth, the sample size matters relative to the population you are studying and what data you are collecting. 1000 Chinese participants in a study on Chinese people is not equivalent in representativeness to a 1000 trans women in a sample of trans women in New York City. A 5 minute survey given to 1000 people is not superior to a case study on 5 people.
This sucks because there's no real rules of thumb. A few years ago, I might've told you to make sure the study has at least 200 people, but that's not necessarily the case since studies can have all different kinds of designs with different sample size requirements. So what should you do? Embrace that a single study is nothing more than that - a single study. Important decisions should never be made based off of single studies, instead, people should look towards meta-analyses and systematic reviews of literature on topics.
A meta-analysis is basically like a study of studies, where the results of many studies looking at the exact same thing are punched in together to see if there truly is an overall effect. They're far from perfect, and meta-analyses have increased scrutiny on their methods today, but they usually help with issues of low sample sizes for individual studies and makes for more representative, actionable conclusions. Systematic reviews involve someone super smart in a particular field interpreting the results of many, different studies to identify patterns and extract some truth about the topics that the field is interested in. These are usually more accessible to read for non-scientists, but only come along every few years because of the gargantuan effort that is required to make them and the time needed for new studies to be done and worth summarizing.
The last thing might just be that scientists care. There's always bad apples, but so many scientists sacrifice secure income throughout their early adulthood, more lucrative job prospects, jobs that require much less stress and work, their mental health (risk of mental illness is much higher than average population due to that stress), their free time to peer review and volunteer work, relationships with partners/family/friends - all just to try and get us a little bit closer to the truth. They pursue truth through times when science gets heavily politicized and sometimes when nobody cares about the specific thing they're super interested in at the moment. They pursue truth knowing they will never get there but it might help someone, somewhere, and in some minuscule part, along the way if they do their part. So, just think twice before you think that a study is ill-intentioned. I assure you, not many would suffer through the bullshit if they didn't care.