Separating multiple processes in implicit social cognition: The quad model of implicit task performance. Conrey, F. R., Sherman, J. W., Gawronski, B.,

Size: px

Start display at page:

Download "Separating multiple processes in implicit social cognition: The quad model of implicit task performance. Conrey, F. R., Sherman, J. W., Gawronski, B.,"

このかひめい
5 years ago
Views:

1 Separating multiple processes in implicit social cognition: The quad model of implicit task performance. Conrey, F. R., Sherman, J. W., Gawronski, B., Hugenberg, K., & Groom, C. J. (2005). Journal of Personality and Social Psychology, 89, Schneider Shiffrin (automatic process) (controlled process) 2 IAT (implicit association test; Greenwald, McGhee, & Schwartz, 1998) MRS (modern racism scale; McConahay, 1986) 1 process pure IAT (Brendl et al., 2001; Mierke & Klauer, 2003) Multiple Automatic and Controlled Processes 1 2 Conceptualizations of Control discriminability (D) overcoming bias (OB) D (accuracy assessment) (e.g., Petty & Wegener, 1999) D D (e.g., Brewer, 1988; Fiske & Neuberg, 1990) OB (self-regulation) Wegner (1994) OB rep: mailto: murakou@orion.ocn.ne.jp 1

2 (D) (OB; cf. Correll et al., 2002; Payne, 2001) Conceptualizations of Automaticity association activation (AC) guessing (G) AC (interference) Schneider & Shiffrin (1977) IAT G 1 (response bias) AC G 2 (AC) (G) Multiple Processes in Implicit Measures 4 IAT 3 (AC) compatible block; incompatible block; (D) (OB) (G) IAT 4 4 quadruple process model (quad model) 1 (familiarity) (e.g., Jacoby, 1991; Yonelinas, G (G) (e.g., Yonelinas & Jacoby, 1994), 3 IAT Figure 3 IAT 2

3 Measuring Multiple Processes: Process Dissociation quad model Jacoby (1991) (process dissociation procedure) quad model The C-First Model of Process Dissociation Jacoby C-First 2 A B 2 (compatible trials) old (incompatible trials) B old A new (C; D ) (A; G 4 ) Figure 1 +, A 5 (e.g., Toth et al., 1994) (AC) The A-First Model of Process Dissociation Lindsay & Jacoby (1994) (Figure 2) (A) (C) A (OB) AC 0 (AC ) 4 G (Buchner et al., 1995; JEP-G) C A Buchner G Figure 1 1 A G 1 G Yonelinas (Yonelinas et al., 1995; M & C) G (guessing) guessing 5 A p(old compatible trials) = C + A(1 C) (1) p(old incompatible trials) = A(1 C) (2) 2 2 A C 0 1 (Batchelder & Riefer, 1990) 3

4 The Quad Model quad model multinominal model (Batchelder & Riefer, 1999; Klauer & Wegener, 1998) 6 AC, D, OB, G 4 IAT Figure 3 AC (D) OB OB AC AC (G) Analyzing Data With the Quad Model quad model (error rate) IAT 7 IAT quad model IAT (P) (N) (W) (B) = 8 G p(correct P, compatible) = AC W P + (1 AC W P ) D P N + (1 AC W P ) (1 D P N ) G (3) p(correct W, compatible) = AC W P + (1 AC W P ) D W B + (1 AC W P ) (1 D W B ) G (4) p(correct N, compatible) = AC B N + (1 AC B N ) D P N + (1 AC B N ) (1 D P N ) G (5) p(correct B, compatible) = AC B N + (1 AC B N ) D W B + (1 AC B N ) (1 D W B ) G (6) p(correct P, incompatible) = AC W P D P N OB attribute + (1 AC W N ) D P N + (1 AC W P ) (1 D P N ) G (7) p(correct W, incompatible) = AC W P D W B OB category + (1 AC W N ) D W B + (1 AC W P ) (1 D W B ) G (8) p(correct N, incompatible) = AC B N D P N OB attribute + (1 AC B N ) D P N + (1 AC B N ) (1 D P N ) G (9) p(correct B, incompatible) = AC B N D W B OB category + (1 AC B N ) D W B + (1 AC B N ) (1 D W B ) G (10) 6 Jacoby multinominal model (Yonelinas & Jacoby, 1996) multinominal model high-threshold yes IAT multinominal model 7 appendix Appendix notation 4

5 AC W P AC B N D P N D W B OB category OB attribute OB category OB OB attribute OB AC B N D B W AC D AC B W D P N AC D OB category > OB attribute 2 D B W D P N D OB D B W D P N AC W P AC B N D OB attribute OB category G χ 2 Sherman HP ( 0 A 0 A 0 χ 2 Testing the Quad Model IAT 1 4 sequential priming paradigm 5 (stochastic validity) Study1 flower-insects IAT 8 (1 AC) D + (1 AC) (1 D) G 1 AC 5

6 Method Results & Discussion 1 χ 2 = 1.74, p = : Parameter Estimates for Flowers-Insects IAT, Study1 Parameter Type Estimate AC insect-unplesant 0.13 flower-pleasant 0.38 OB attribute judgment 0.77 category judgment 0.87 D 0.88 G 0.50 AC D OB category 0 G 0.5 OB attribute 0 (χ 2 (1) = 2.39, p = 0.12) AC D OB IAT Study (OB, D) (AC, G) Method Sherman HP 4% 10% 10 (functional independence) 6

7 IAT ms- 670ms Results & Discussion 6% 14% 2 AC G χ 5 = 3.19, p = : Parameter Estimates for Black-White IATs, Study2 Parameter Comparison Estimate AC Black-unplesant 0.11 White-pleasant 0.18 G practice 0.52 IAT RW-IAT OB attributes names D G test blocks D OB AC AC G Study 3 3 G G Method

8 2 Results & Discussion 5% AC, D, OB χ 9 = 12.21, p = : Parameter Estimates for Flowers-Insects IAT, Study3 Parameter Comparison Estimate AC insect-unplesant 0.11 flower-pleasant 0.11 OB attributes 0.00 names 0.53 D 0.95 more right-handitems more left-hand items G G G (χ 2 (1) = 0.84, p = 0.36) AC OB category OB attribute Study 4 4 IAT 4 4 IAT AC IAT OB IAT 11 D G Method 24 IAT 2 Results & Discussion 6% (χ 2 (2) = 3.26, p = 0.20) OB IAT OB 8

9 4: Parameter Estimates for the Black-White IAT, Study4 Parameter Comparison Estimate AC Black-unplesant 0.04 White-pleasant 0.09 OB attributes 0.00 names 0.52 D 0.93 G 0.54 IAT (Greenwald et al., 1998) 3000ms (0.7%) 3000ms 300ms (0.1%) 300ms (Greenwald et al., 2003) quad model 5 IAT 6 6 5: Parameter Estimates for the Black-White IAT, Study4 Parameter Comparison Estimate M S D AC Black-unplesant White-pleasant OB attributes names D G : Multiple Regression IAT Score, Study4 IAT old algorithm IAT new algorithm Term Comparison β p β p AC Black-unpleasant White-pleasant OB attributes names D G D IAT 9

10 Study 5 5 quad model Payne (2001) sequential priming 2 1 C-first model A-first model quad model 1 (social accountability) (Lambert et al., 1996) Lambert et al.(2003) Lambert et al. (2003) C-first model A C Lambert C-first model Lambert quad model Method ms 200ms 100ms or 450ms Results & Discussion AC Whitemale tool,ac White f emale tool, AC Blackmale gun, AC Black f emale gun, OB, D, G % (χ 2 (3) = 5.17, p = 0.16) C-first model A-first model quad model (χ 2 (6) = 6.71, p = 0.35; χ 2 (6) = 13.25, p = ) 12 A-first model C-first model quad model 10

11 7: Parameter Estimates for Sequential Priming Task Used by Lambert et al. (2003), Study5 Parameter Prime Estimate public context private context AC White male-tool White female-tool Black male-gun Black female-gun OB D G 0.46 Lambert D (χ 2 (1) = 51.47, p < 0.001) (χ 2 (2) = 8.24, p = 0.01) (Eagly & Kite, 1987) OB G Lambert OB 2 General Discussion quad model (e.g., Blair et al., 2001; Dasgupta & Greenwald, 2001) (e.g., Devine & Monteith, 1999; Moskowitz et al., 1999) quad model IAT (e.g., Cunningham et al., 2001) 11

P /ﬁ¡ﬁcﬁNŒç

P /ﬁ¡ﬁcﬁNŒç implicit memory levels of processing Craik Lockhart 1972 1988 cf. 1988 1 1980 1980 1 explicit memory implicit memory e.g., Graf & Schacter, 1985; Schacter, 1987 2001 priming direct priming effect 2 1 2