Fascinating preprint, @DHBostyn & @xphilosopher! A few journal club thoughts from me and @AllardThuriot (Thread): 1/n
1) We loved the goal of mathematically modelling different shapes of blame/praise judgments, and identifying neutral points. I saw this as very much in line w/ calls to reduce the arbitrariness of measures and define the continua of constructs, e.g.: psycnet.apa.org/record/2006-00… 2/n
2) We thought the answers to the real-world questions around timeliness of email replies, # of blood donations, time spent comforting a friend, etc. were an important contribution. We probably spent an hour just staring at this graph (Figure 8): 4/n 
3) We also loved the transparent reporting and detective-style narrative (e.g., openly exploring new possibilities instead of HARKing when the original hypotheses didn't work out) 😛 5/n
Here are some questions/suggestions. 1) Could there be floor effects here? It seems implausible that cheating 20 times is no worse than cheating 2-3 times. "Extremely blameworthy" might include a larger range than what the scale points allow participants to express. 6/n
2) It would have been nice to see more discussion of the distinction between supererogatory vs. duty-based behaviors (or perfect vs. imperfect duties; e.g., Trafimow & Trafimow, 1999), and how they relate to parameters of the curve. 7/n
Almost by definition, supererogatory behaviors should be defined by having neutral points at 0 occurrences of the behavior (i.e., an intercept of 0). And that is what the data seem to suggest: 8/n
So even though the shape is concave for some duty-based behaviors and superogatory behaviors, the intercept seems to be meaningfully different. 9/n
3) We found the "blarging" studies more difficult to interpret. It is hard to imagine what participants were thinking when making these decontextualized judgments. See, e.g.: journals.sagepub.com/doi/abs/10.117… @ChelseaSchein 10/n
The # of Ps who learn that blarging is bad but still rate it as praiseworthy (yellow triangles) implies that many potentially did not understand the task. 11/n
Also, because frequency was the only information in the "blarging" studies, it makes sense that participants can only use frequency info (in the absence of other context). In real-world scenarios, other information might completely trump frequency. 12/n
4) Finally, we loved the discussion of potential individual differences in the blame/praise curves. We were curious about how different the results might have been using a within-person design (e.g., asking participants to draw their own curves). 13/n
Overall, this was an incredibly stimulating and thought-provoking paper. We hope that future researchers will continue to model these kinds of nonlinear judgment processes! 14/14
