Skip to main content
. 2015 Aug 4;112(33):E4512–E4521. doi: 10.1073/pnas.1419828112

Table 1.

Postsearch shifts in voting preferences for study 1

Experiment Candidate Rating Mean deviation from control (SE)
Gillard bias u Abbott bias u
1 Gillard Impression 1.44 (0.56)* 761.0 −1.52 (0.56)** 380.5
Trust 1.26 (0.53)** 779.0 −1.85 (0.48)** 330.5
Like 0.26 (0.54) 615.5 −1.73 (0.65)** 387.0
Abbott Impression −2.29 (0.73)** 373.0 1.11 (0.72)** 766.5
Trust −2.02 (0.63)** 384.0 0.67 (0.76) 679.0
Like −1.55 (0.71) 460.5 1.17 (0.64)* 733.0
2 Gillard Impression 0.97 (0.65) 704.0 −2.38 (0.79)*** 325.0
Trust 0.94 (0.72) 691.5 −2.17 (0.74)** 332.5
Like 0.55 (0.76) 639.5 −1.82 (0.66)** 378.0
Abbott Impression −1.44 (0.81)* 395.5 1.17 (0.75)* 742.0
Trust −0.79 (0.81) 453.5 1.85 (0.72)** 774.5
Like −1.44 (0.70)* 429.0 0.64 (0.71) 690.0
3 Gillard Impression 1.44 (0.73)* 717.5 −0.55 (0.69) 507.5
Trust 0.47 (0.70) 620.0 −0.23 (0.56) 466.5
Like 0.44 (0.65) 623.5 −0.41 (0.70) 528.5
Abbott Impression −0.32 (0.70) 534.0 1.26 (0.60)* 750.5
Trust −0.73 (0.65) 498.5 1.50 (0.58)** 795.0
Like −0.50 (0.61) 496.0 0.88 (0.62) 681.5
*

P < 0.05, **P < 0.01, and ***P < 0.001: Mann–Whitney u tests were conducted between the control group and each of the bias groups.