Mindfulness reduces the correspondence bias

ABSTRACT

The correspondence bias (CB) refers to the idea that people sometimes give undue weight to dispositional rather than situational factors when explaining behaviours and attitudes. Three experiments examined whether mindfulness, a non-judgmental focus on the present moment, could reduce the CB. Participants engaged in a brief mindfulness exercise (the raisin task), a control task, or an attention to detail task before completing a typical CB measure involving an attitude-attribution paradigm. The results indicated that participants in the mindfulness condition experienced a significant reduction in the CB compared to participants in the control or attention to detail conditions. These results suggest that mindfulness training can play a unique role in reducing social biases related to person perception.

KEYWORDS:

• Correspondence bias
• Mindfulness
• Fundamental attribution error

Notes

1. A 2 (condition: mindfulness vs. control) × 2 (essay position: for vs. against nuclear power) analysis of variance (ANOVA) was conducted on the degree to which participants considered the writer to be for or against the use of nuclear power. The main effect of mindfulness condition was not significant, F < 1. The expected main effect of essay position was significant, F(1, 87) = 237.56, p < .001, η² = .73. This effect indicated that participants who read an essay opposed to nuclear power considered the writer to be more strongly opposed to nuclear power (M = 1.94, SD = 1.39) than did participants who read a pro-nuclear power essay (M = 6.04, SD = 1.17). There was a marginally significant interaction between condition and essay position, F(1, 87) = 3.61, p = .06, η² = .04.

2. We conducted a 2 (condition: mindfulness vs. control) × 2 (essay position: for vs. against nuclear power) ANOVA without the covariate on the degree to which participants considered the writer to be for or against the use of nuclear power. The main effect of mindfulness condition was not significant, F < 1. The expected main effect of essay position was significant, F(1, 104) = 168.99, p < .001, η² = .62. This effect indicated that participants who read an essay opposed to nuclear power considered the writer to be more strongly opposed to nuclear power (M = 1.98, SD = 1.50) than participants who read a pro-nuclear power essay (M = 5.72, SD = 1.53). There was a significant interaction between condition and essay position, F(1, 104) = 4.79, p < .05, η² = .04.

Further analysis revealed that in the “for” essay condition, the difference between the mindfulness (M = 5.33, SD = 1.69) and control (M = 6.11, SD = 1.28) conditions approached significance, F(1, 104) = 3.65, p = .06, η² = .03. In the “against” essay condition there was no significant difference between the mindfulness (M = 2.22, SD = 1.85) and control (M = 1.74, SD = 1.02) conditions, F(1, 104) = 1.40, p = .06, η² = .03. In the mindfulness condition, the difference between the “for” (M = 5.33, SD = 1.69) and “against” (M = 2.22, SD = 1.85) essay positions was significant, F(1, 104) = 58.44, p < .001, η² = .36. This was also the case for the control condition, F(1, 104) = 115.33, p < .001, η² = .53 (M_for = 6.11, SD = 1.28, and M_against = 1.74, SD = 1.02).

3. We conducted a 3 (condition: mindfulness vs. control vs. attention to detail) × 2 (essay position: for vs. against nuclear power) ANOVA without the covariate on the degree to which participants considered the writer to be for or against the use of nuclear power. The main effect of mindfulness condition was not significant, F(2, 163) = 1.37, p = .26, η² = .02. The expected main effect of essay position was significant, F(1, 163) = 406.79, p < .001, η² = .71. This effect indicated that participants who read an essay opposed to nuclear power considered the writer to be more strongly opposed to nuclear power (M = 1.75, SD = 1.32) than did participants who read a pro-nuclear power essay (M = 5.86, SD = 1.36). The interaction between condition and essay position was not significant, F(2, 163) = 2.18, p = .12, η² = .03.

Additional information

Funding

This work was supported by the Leverhulme Trust [grant number F/00 236/AB to the first, fourth, and fifth authors].