3 ms·

If it's not bullshit, then I definitely need a ELI5 explanation of this: https://imgur.com/38CXnRx https://imgur.com/38CXnRx

• frickinLasers 7y ago
  This is my interpretation as someone who has never seen this type of study and is giving the authors the benefit of the doubt: Take Table 3. The outcome "O_ij" for each student's "views of abilities" is the scale score (split into self-efficacy and problem solving). The table shows bullshitters think they're awesome, giving positive scores (higher than the mean). Bullshitters were previously categorized into four groups by asking questions about what math terms they were familiar with, including terms that don't exist (maybe they were familiar with 0,1,2, or 3 of the fake terms) Control for other factors by adding levers for achievement, socio-economic status, gender and school. These are the capital letters, set by test scores, etc. Students from each school will have a different u_j, probably higher for higher-performing schools B.S. is probably a matrix of ones, or 1-4, or something. Perform a regression (let a computer do some matrix math to solve for unknown variables), using all of the students as data points. Each student will end up with a different set of italic letters indicating how much the capital letters influence their outcome, with students who are similar in one aspect sharing similar (or equal, depending on how they set up the problem) values for that italicized letter. For table 3, it looks like the study claims students who have been identified as bullshitters end up with higher beta values ("regression results"). It's possible this explanation is bullshit, though.