Ra#onal Order Effects in Responsibility A6ribu#ons M D T Tobias Gerstenberg, David A. Lagnado, Maarten Speekenbrink & Catherine Loz Sze Cheung C contact: t.gerstenberg@ucl.ac.uk Experiment Who is the cause? Who is responsible? D T M C Responsibility A6ribu#on Phase Probability Ra#ng Phase Non-­‐iden#cal Scores recency? Result Outcome Sum of 1st & 2nd 3 Low Score High Score primacy? 1) nature of the chain (Miller & Gunasegaram, 1990) à events causally dependent or not Matching Pennies Game 2) “logic-­‐based accounts” (Mandel, 2003) 1st à necessary vs. sufficient condi#ons Result Outcome 1st player 2nd player 3rd player 92% à degree of change of the probability of the effect Matching Pennies Game IniDal 1st Event 2nd Event P(Win) .5 .5 1 or 0 1000 Mul#ple Choice Game IniDal 1st Event 2nd Event .25 0 0 P(Win) .25 .5 0 .25 .5 1 A B C D 2nd A B both performance and certainty of the team’s outcome influence blame/credit aIribuDons Problems with the probabilistic account 1) nature of the events … 1 … 1 2 3 2 3 4 3) quality of performance * certain uncertain * * Probability Ra#ng Phase 2) recency 8 4 knowledge condi#on (Experiment 1, N = 41) no knowledge condi#on (Experiment 2, N = 56) * -­‐ performance effect (Experiment 1 & 2) -­‐ certainty of outcome effect (Experiment 1) -­‐ performance effect even if the result is already determined? 1000 between-­‐subject varia#on: loss win certain uncertain uncertain certain 1 2 3 4 6 7 8 9 1 2 3 4 6 7 8 9 1 2 3 4 6 7 8 9 Predictions & Results Mul#ple Choice Game 1st win uncertain certain 11 12 13 16 17 18 4 3 2 4 3 2 8 7 6 8 7 6 Iden#cal Scores 2nd 3) probabilis#c account (Spellman, 1997) loss certain uncertain 2 3 4 6 7 8 4 3 2 4 3 2 8 7 6 8 7 6 4 vs. 2 4 Responsibility A6ribu#on Phase Responsibility A6ribu#on Phase Discussion Why raDonal? à What can be inferred from an agent’s performance in a certain outcome situaDon? -­‐ high score: agent is in principle capable of performing well -­‐ low score: ambiguous evidence à did not try hard vs. is not able Research Question How do (i) quality of performance and (ii) the extent to which a contribu#on made a difference to the result affect people’s percepDon of how responsible each contribuDon was for the eventual outcome? Why should it maIer, it doesn’t make a difference?! -­‐ not in the actual world but in many possible worlds it would have made a difference References Gerstenberg, T., Lagnado, D. A, Speekenbrink, M. & Cheung, C. (2011). RaDonal order effects in responsibility aIribuDons. In L. Carlson, C. Hölscher, & T. Shipley (Eds.), Proceedings of the 33rd Annual Conference of the Cogni5ve Science Society. AusDn, TX: CogniDve Science Society, 1715-­‐1720. Spellman, B. A. (1997). CrediDng causality. Journal of Experimental Psychology: General, 126 (4), 323-­‐348.