Rationalism can have two definitions: first, a community that call themselves Rationalists, that read ‘high-IQ sites’ such as Marginal Revolution, Less Wrong, and Slate Star Codex, and according to various surveys, identify as liberal, are atheist or agnostic, and, in general, hold a ‘realist’ philosophical worldview.
Second, an epistemological approach–to try to see the hidden side of everything (like the Freakonomics apple that has an orange interior) as way to attain a higher understanding, instead of just taking things for granted or prima facie. In the spirit of Michael Shermer and the ‘skeptic movement’, rationalism involves debunking sensationalism, fallacious reasoning, and partisanship that may obscure the underlying truth/reality. Unlike the first definition, the second doesn’t imply or connote any specific political leaning, and such a definition can be applicable to anyone, including even reactionaries, so although the first and second definitions may be correlated, one need not be a liberal atheist to apply rationalism.
Philosophically, rationalists–in the context of the Rationalist community–tend to reject absolutism. Empiricism (such as behavior studies, drug studies, and psychology) and inductive reasoning play heavily, and also to some extent, Platonism (universal ‘truths’ such has Newton’s laws of motion, that exist independent of observation), but rationalists tend to reject rationalism in the philosophical sense. In terms of ontology, rationalists tend to be positivists and materialists. The confusion is that rationalists (in the context of the Rationalist community), in terms of epistemology, are empiricists, not rationalists.  Also, economically, to be ‘rational’ means maximizing one’s expected value, among all possible choices.
Here are some examples of the second definition as applied to specific news stories and ‘thought experiments’:
A couple day ago, a story on Bloomberg This Is Why Australia Hasn’t Had a Recession in Over 25 Years went hugely viral, garnering hundreds of votes and comments on social news sites such as Reddit and Hacker News. On the surface, it seemed like Australia had found the ‘holy grail’ to eternal economic growth–or so it seemed–but was there some key detail omitted in the reporting of this so-called Australian ‘miracle’, that few people noticed and was otherwise buried beneath the hype? Upon seeing the headline, it took me no more than two minutes to debunk the premise of the article, with the help of some quick Googling. Australia’s 25-year stretch of uninterrupted growth is in its ‘local currency’ but when converted into US dollars, shows multiple recessions: in the late 90’s to 2002, from 2008-2009, and since 2013:
So what gives? Why do we need to convert Australia’s growth to US dollars? Anyone who studied even econ 101 knows that nominal figures are meaningless–they need to be put in the context of something–either against inflation or a benchmark currency. Because the United States and the US dollar are the biggest and most important economy and currency, and because commodity prices and GDP growth are denominated in US dollars, it’s a benchmark that everyone can agree upon, and that’s why the US dollar is used as a benchmark for ranking the GDP growth of foreign countries, not local currencies. It’s impossible to have a ranking system where each country uses their own currency. Without such a conversion, the global leader of economic growth and stock market gains would be…Zimbabwe, with its 10,000% annual inflation and toilet paper currency:
Wow, Zimbabwe’s economy must sure be booming. Bloomberg should write an article about that. If they do, you’ll know the backstory.
A Bridge Too Far
For the second example, the I-85 bridge collapse in Atlanta on March 30th, 2017, which inconvenienced 250,000 commuters and made global headlines–the collapse even has it own Wikipedia page. Some even saw the I-85 collapse as a sort of metaphor or harbinger for the collapse of Western civilization. But what was ignored or overlooked by the media firestorm are the thousands of bridges that don’t collapse (the United States has 607,380 bridges, according to the federal National Bridge Inventory)…the fact this is such big news speaks to its rarity. Given all the thousands of bridges and the millions of vehicles that cross them everyday, such events are statistically inevitable, and it’s almost a miracle such collapses aren’t more common. That’s not to say America’s infrastructure cannot benefit from improvement, but rationalism is about putting things in perspective, using data to try to have ‘correct‘ opinions, instead of being swayed by media narratives.
For the third example, wisdom teeth extraction is touted as a safe procedure. However, complications such as severe bleeding, pain, and nerve damage may occur days or even week after the procedure, but such parents often go to the hospital instead of the dentist, so such complications may be recorded as being isolated instead of accompanying the extraction; thus, only counting in-patient complications overestimates the safety of wisdom teeth extraction. This is just a hypothetical example, but shows the thought process of rationalism, in which omitted information (out-patient complications) can result in misleading conclusions.
An finally, number four, why you shouldn’t trust successful people’s advice, due to survivorship bias. Rationalism tries to understand why people are susceptible to cognitive and logical biases and fallacies (for example, the ‘recency bias/effect’ regarding coverage of the I-85 collapse, and the ‘fallacy of composition’ regarding the US economy and how so many people got the post-2009 economic recovery wrong), and how one can avoid being taken by appealing but fallacious narratives.
 Less Wrong is closer to being rationalist, in the philosophical sense of the word. Consider the the Monty Hall game, in which Bayes Theorem (an abstract concept) is invoked to prove that switching is optimal. This is related to Platonic realism (math concepts). But an empirical approach to the Monty Hall problem is also possible by running large simulations, but the application of probability theory (rationalism) bypasses this.