Fwd: "Life After P-Hacking" & "Six Biases That Are All The Same", Special Cognition Seminar, Thursday March 24th: Dr. Joseph Simmons (Guest Speaker)
Danniel Varona-Marin
From: Evan Risko <efr...@uwaterloo.ca>
Date: Fri, 18 Mar 2016 at 10:08
Subject: "Life After P-Hacking" & "Six Biases That Are All The Same", Special Cognition Seminar, Thursday March 24th: Dr. Joseph Simmons (Guest Speaker)
To: Evan Risko <efr...@uwaterloo.ca>, psyc...@lists.uwaterloo.ca <'psyc...@lists.uwaterloo.ca'>, psyc...@lists.uwaterloo.ca <'psyc...@lists.uwaterloo.ca'>, Catherine Burns <catheri...@uwaterloo.ca>, Edith Law <edit...@uwaterloo.ca>, danni...@gmail.com <'danni...@gmail.com'>, Jessie Chin <jessi...@uwaterloo.ca>
Greetings,
Next week the Cognitive area is very excited to present a guest speaker, Dr. Joseph Simmons (Wharton School of the University of Pennsylvania). Dr. Simmons is a leader in both recent discussions about p-hacking and research on judgment and decision-making. He will be giving two talks “Life After P-Hacking” and “Six Biases That Are All The Same” (abstracts below) on Thursday March 24th at 11:30 and 2:30 respectively. Both talks will take place in the PAS building, room 3026 (ABC Room). For more information about Dr. Simmons and his work please see his website (https://oid.wharton.upenn.edu/profile/1666/research), blog (http://datacolada.org/) or follow him on twitter (https://twitter.com/jpsimmon?lang=en).
Thursday March 24th, 11:30 am: Life After P-Hacking
P-hacking is the practice of conducting many analyses on the same dataset until one achieves a reportable, statistically significant result (p < .05). P-hacking can lead researchers to believe in, and publish, findings that are false (i.e., not replicable), and it is likely to have played a major role in psychology’s replicability crisis. Since p-hacking is so bad, we should work hard to stop doing it. In this talk, I discuss how a decision to stop p-hacking affects how research is done and judged.
Thursday March 24th, 2:30 pm: Six Biases That Are All The Same
Investors decide whether Apple’s stock price is too high. Gamblers decide if Toronto’s win probability is greater than 45%. Consumers decide whether an apartment is worth more than its monthly rent. Participants decide whether they are more intelligent than the average student. Despite differences in content, these decisions all require the same task: deciding whether some entity is greater or less than another entity. Such over/under decisions are ubiquitous.
We propose that over/under decisions are heavily influenced by spontaneous evaluations. When deciding whether A is greater than B, people first spontaneously evaluate and compare them. Biases emerge from the fact that people do not spontaneously evaluate moderate or unknown quantities, such as well-calibrated point spreads, apartment rents, or others’ average intelligence, leading them to make spontaneous decisions between A and B based primarily on their evaluation of A.
Data from online casinos (N=1,007 games) and five experiments (Ns between 304-510) reveal six different biases produced by this process, including sports bettors’ preference for favorites over underdogs, sports bettors’ tendency to bet that high-scoring games will score more than oddsmakers predict, people’s tendency to predict that hot (cold) cities will be hotter (colder) than usual, and people’s tendency to predict well-performing (poorly performing) companies’ stocks to go up (down). We also show that the above- and below-average effect and the prominence effect are explained by this process as well.
Have a nice day,
Evan
------------------------------------
Evan F. Risko
Assistant Professor
Canada Research Chair
Cognitive Area Head
Department of Psychology
University of Waterloo