copied from twitter / reddit
Archive link: Attention Required!
That's how true science is done today: you find sources that confirm your biases, and ignore the ones disagreeing with them.
Archive link: Attention Required!
A Reddit user summarizing said:Lee Jussim, Chair of the Department of Psychology at Rutgers, chronicles some examples of leftist bias in psychology research. See his tweets for details. I'll summarize the absolutely shocking key points (skip to #5, #6, and #7 for the worst):
- A 2018 study uncovered 17 UNPUBLISHED national surveys with representative samples totalling >10,000 respondents finding (1) no evidence of racial bias among whites and (2) "a small-to-moderate net discrimination in favour of black targets among blacks."
- A 2009 study responding to criticisms of the IAT (the implicit association test; "implicit bias") which "reported 10 studies no researcher should ignore" on the supposed reality of implicit bias. However, 7 out of the 10 papers did not even address racial bias. Of the 3 papers that did, they reported 4 studies on racial bias. 2 found no evidence of racial bias at all. 1 found a small bias but failed to test whether it "statistically differed from egalitarian responding." 1 found bias in 1/3 of conditions (no bias in 3/4 of conditions).
- One of the "most famous papers in early social psych" "claimed to show there is no reality, everything is subjective perception." See his tweet for the details. Essentially, it found that 92% of the time, there was no bias in perceptions at all. The study ignored this fact and based its extreme conclusion (everything is subjective) on the remaining 8%.
- A 2018 "review in the prestigious and high impact Annual Review of Psychology concluded that gender stereotypes were mostly inaccurate." However it ignored "11 papers reporting 16 studies that assessed gender stereotype accuracy." What did those 11 papers find? That gender stereotypes are (usually) more valid than most social psychological hypotheses.
- A 2012 paper "showing pro-male gender bias in science fields hiring a lab manager." Three years later, in 2015, a paper was published "showing pro-female gender bias across STEM fields in faculty hiring." Only counting AFTER the 2015 study came out, the 2012 paper has been cited over 7x more often. In other words, well over 1000 papers cite the 2012 paper without mentioning the 2015 one. "What's even more extraordinary is the superiority of the 2015 study on virtually all quantifiable measures of methods quality."
- "Routinely, studies finding pro-male bias in peer review are: 1. lower methodological quality 2. More highly cited than studies finding pro-female bias." Jussim reviewed the literature and found "far smaller samples but higher citations for studies finding pro-male bias." The mean sample size for studies finding pro-male bias was 825.5; for studies finding pro-female bias the sample was 11,385.67. Yet, mean citations of the pro-male studies were 91.75 and for the pro-female studies only 26.83.
- A 2003 meta-analysis showed "conservatives WAY MORE rigid/dogmatic than liberals." A 2010 meta-analysis "found mixed/weak evidence of such lib/con differences." The 2010 study had 127 studies and was cited 173 times since 2011. The 2003 study only had 88 studies, yet it was cited 3060 times since 2011.
That's how true science is done today: you find sources that confirm your biases, and ignore the ones disagreeing with them.