Ioannis Evangelidis
@ievangelidis.bsky.social
Associate professor at ESADE. Behavioral scientist interested in judgment & decision-making, measurement, data analytics, replications, open science, etc.
Never thought about it. You describe inferences about a magnitude based on a categorical variable. In the paper we study inferences about one magnitude based on another (inferences about liking differences based on consensus). Perhaps there is less ambiguity, but inferential processes may be similar
February 3, 2025 at 1:19 PM
Never thought about it. You describe inferences about a magnitude based on a categorical variable. In the paper we study inferences about one magnitude based on another (inferences about liking differences based on consensus). Perhaps there is less ambiguity, but inferential processes may be similar
Full paper (open access): academic.oup.com/jcr/advance-...
P.S. This is Graham's dissertation paper during his PhD at Bocconi, supervised by Joachim Vosgerau and myself. It's a special one.
P.S. This is Graham's dissertation paper during his PhD at Bocconi, supervised by Joachim Vosgerau and myself. It's a special one.
People Believe If 90% Prefer A over B, A Must Be Much Better than B. Are They Wrong?
Abstract. We show that consumers confuse consensus information in polls—such as 90% prefer product A over product B—with differences in liking—the extent t
academic.oup.com
February 3, 2025 at 9:24 AM
Full paper (open access): academic.oup.com/jcr/advance-...
P.S. This is Graham's dissertation paper during his PhD at Bocconi, supervised by Joachim Vosgerau and myself. It's a special one.
P.S. This is Graham's dissertation paper during his PhD at Bocconi, supervised by Joachim Vosgerau and myself. It's a special one.
Not a review paper but this one may be useful. It reexamines some conclusions from previous research doi.org/10.1037/0096...
APA PsycNet
doi.org
December 11, 2024 at 9:46 PM
Not a review paper but this one may be useful. It reexamines some conclusions from previous research doi.org/10.1037/0096...
If you want estimates of main effects either use a different coding (-1 and 1 instead of 0 and 1) or run a regression model without interaction term(s). This should be common knowledge by now, but I keep encountering this issue in papers that I review and sometimes even in accepted papers.
November 20, 2024 at 1:06 PM
If you want estimates of main effects either use a different coding (-1 and 1 instead of 0 and 1) or run a regression model without interaction term(s). This should be common knowledge by now, but I keep encountering this issue in papers that I review and sometimes even in accepted papers.
Not trying to defend Wansink, but this feature of the study should influence the magnitude of the effect size. It is an important mechanical aspect that is not really discussed anywhere in the replication 3/3
November 9, 2023 at 11:52 PM
Not trying to defend Wansink, but this feature of the study should influence the magnitude of the effect size. It is an important mechanical aspect that is not really discussed anywhere in the replication 3/3
This is consistent with the fact that the difference in mean ounces of soup consumed between original and replication is larger in the treatment group than in the control group 2/3
November 9, 2023 at 11:52 PM
This is consistent with the fact that the difference in mean ounces of soup consumed between original and replication is larger in the treatment group than in the control group 2/3
Cool findings. Congrats!
October 24, 2023 at 9:35 AM
Cool findings. Congrats!