Personality Traits Moderate Attribute Framing Effects

Abstract

Previous research on attribute framing effects focused on context-specific variables that moderate it. This research examined whether two personality traits, namely agreeableness and conscientiousness, moderate the effect of attribute framing on the perceived fairness of allocation criteria. Two experiments showed that attribute framing affected the perceived fairness of allocation criteria for participants who score high on these personality traits. In contrast, participants who score low on these personality traits were relatively immune to attribute framing effect. Critically, these personality traits did not moderate the effect of attribute framing on evaluations in a nondistributive justice scenario. These findings are consistent with the possibility that highly agreeable and highly conscientious people are more sensitive to social issues, and, consequently, are more susceptible to attribute framing in scenarios that involve distributive justice. Theoretical implications are discussed and future research is suggested.

Keywords

attribute framing agreeableness conscientiousness social justice

Introduction

Since Tversky and Kahneman (1981) introduced the systematic reversals of preferences that people display in alternate framings of problems, contingencies, or outcomes, numerous studies examined the effect of framing in various contexts. While many of these studies focus on the situational factors that affect such biases, some studies suggest that individual differences may also affect such biases. The present article investigates whether personality traits moderate the effect of attribute framing in contexts that do or do not involve social justice.

Attribute Framing

Levin, Schneider, and Gaeth (1998) proposed that framing studies could be divided into three major distinct types: risky choice framing, goal framing, and attribute framing. Levin et al. showed how these three types of framing effects are distinguishable in terms of their operational definitions, their typical results, and the likely underlying processes. This article focuses on attribute framing in which the description of a critical attribute of an object or an event is presented in either positive or negative framing. Although the very same object is presented using different descriptions, attribute framing has been shown to bias the evaluation of the relevant object. For example, Levin and Gaeth (1988) showed that perceptions of the quality of ground beef were more favorable when the beef was labeled as “75% lean” (positive framing) relative to “25% fat” (negative framing). The attribute framing bias is a robust effect that was consistently found in different contexts. Summarizing 36 attribute framing studies, Levin et al. concluded that in most studies positive framing results in more favorable evaluations of objects or events relative to negative framing. Furthermore, in a meta-analysis of 30 studies, Piñon and Gambara (2005) reported an average attribute framing effect size (Cohen’s d; Cohen, 1988) of 0.26. Levin et al. proposed a theoretical explanation for attribute framing bias that focused on the encoding of information in memory: “The positive labeling of an attribute leads to an encoding of the information that tends to evoke favorable associations in memory, whereas the negative labeling of the same attribute is likely to cause an encoding that evokes unfavorable associations” (p. 164).

Several studies examined the potential moderators of attribute framing effects. For example, women were found to be more susceptible to attribute framing effect when manipulating percentage of fat in chocolate (as either 20% fat or 80% fat free) relative to men (Braun, Gaeth, & Levin, 1997). The authors explained the moderation of gender by claiming that women pay more attention to the fat content than men, and hence, they were affected by the attribute framing more than men. In yet another context, Republicans opposed a mandatory carbon fee more when it was labeled as a carbon tax than when it was labeled as a carbon offset. In contrast, Democrats’ attitudes were unaffected by the tax versus offset labeling, and Independents’ attitudes were affected by the attribute framing, although not as much as the Republicans (Hardisty, Johnson, & Weber, 2010). The authors explained the moderation of political party self-identification by claiming that Republicans and Independents oppose taxes more than Democrats, and hence, they were more adversely affected by the negative framing of taxes than the Democrats.

These studies suggest that individual differences moderate attribute framing bias, such that people who are more sensitive to an issue are more susceptible to attribute framing effect, while the complementary groups are relatively immune to it.

The Effect of Attribute Framing on Allocation Criteria

Several studies have shown that people are affected by attribute framing of distributive justice situations when asked to judge the fairness of non-egalitarian allocation criteria (e.g., waiting time, need, effort, and ability; Gamliel & Peer, 2006, 2010). A typical distributive justice scenario involves scarce goods that are desired by many but can be delivered only to some. People typically judge the non-egalitarian criteria to be more fair when they are presented with positive framing as in who should be given the goods, relative to presenting them with the objectively identical information of the complementary framing of who should be denied the goods. For example, when presented with 50 patients who need a heart transplant in order to survive, while only 25 hearts are available, participants perceived allocating the organs to 25 patients who are waiting for the longest period of time for a transplant as more fair than not allocating the organs to 25 patients waiting the shortest time (Gamliel & Peer, 2010).

The Relationship Between Justice Judgments and Personality Traits

Agreeableness and conscientiousness are two of the five factors of the Big Five Factor Model (BFF; Costa & McCrae, 1985). Agreeableness is the tendency to be compassionate and friendly rather than suspicious and antagonistic, with a positive and optimistic view of human nature. Individuals who are highly agreeable are invested in getting along with others. They tend to be considerate, friendly, generous, helpful, and will sacrifice their own interests in the service of others. Conscientiousness includes responsibility, self-discipline, efficiency, acting dutifully, having high standards, and aiming for achievement. Highly conscientious individuals are more controlled, regulated, and less impulsive (Goldberg, 1990). These personality traits seem to relate to social attitudes so that individuals who are high in agreeableness and conscientiousness should be highly invested in being fair and sensitive to matters of justice.

Several studies found a positive association between agreeableness and conscientiousness and different aspects of social justice: ethical leadership (Kalshoven, Den Hartog, & De Hoogh, 2011), justice sensitivity (Schmitt, Baumert, Gollwitzer, & Maes, 2010), job applicants’ judgments of hiring fairness (Truxillo, Bauer, Campion, & Paronto, 2006), egalitarianism, fairness-reciprocity, and in-group loyalty aspects of the moral foundations questionnaire (Hirsh, DeYoung, Xu, & Peterson, 2010), and distributive justice evaluations (Shi, Lin, Wang, & Wang, 2009).

Research Hypotheses

Assuming that individuals who are highly agreeable or highly conscientious are more sensitive to issues of distributive justice, the current research hypothesized that they would also be more susceptible to attribute framing in contexts of distributive justice. Specifically, we examined two hypotheses:

Hypothesis 1: Attribute framing would affect participants’ judgments in the context of distributive justice and in a context that is not related to social justice. Thus, participants would rate the perceived fairness of the effort, waiting time, need, and merit criteria higher when presented with positive framing of distributive justice situations (i.e., who should get the good) relative to negative framing of the same situation (i.e., who should be deprived of the good). Similarly, in a context that is not related to social justice, participants would evaluate an object more favorably when its attribute is presented with a positive framing relative to negative framing of the same attribute.

Hypothesis 2: In the context of distributive justice, participants who would score high on agreeableness or conscientiousness would be more susceptible to attribute framing relative to those who would score low. However, these personality traits would not moderate attribute framing in the context that is unrelated to social justice.

The first hypothesis is a replication of previous findings, while the second hypothesis is a unique and novel contribution of the current research. These hypotheses were examined in two experiments: The first was designed to show that agreeableness and conscientiousness moderate attribute framing in a distributive justice scenario. The second was designed to replicate this finding, to generalize it to other scenarios of distributive justice, and to examine whether both personality traits would moderate attribute framing in a context that is unrelated to social justice.

Experiment 1

Method

Participants

Participants were 117 Israeli adult volunteers aged 25–70 (M = 42.4; SD = 12.3; 64% were women).

Materials

Agreeableness and Conscientiousness Questionnaires

The Agreeableness and Conscientiousness scales of the BF-44 (John, Donahue, & Kentle, 1991) in the Hebrew translation (Etzion & Laski, 1998) included 18 items of the shortened version BFF questionnaire. All participants were presented first with the 9 items asking about agreeableness, followed by the 9 items asking about conscientiousness. All items were on 5-point Likert-type scale, and high values represented high agreeableness and conscientiousness. The internal reliabilities (Cronbach’s α) of the agreeableness and conscientiousness items were .75 and .81, respectively. These reliability measures are somewhat higher than the respective ones reported for the large multisite study (.67 and .77; Schmitt, Allik, McCrae, & Benet-Martínez, 2007).

Distributive Justice Questionnaire

The scenario presented participants with statistical data of child-adoption requests in Israel, where there are more qualified people who apply for adoption than children available for adoption. Hence, the social services have to apply additional criteria in order to select those who will be given children to adopt. Participants were asked to judge the perceived fairness of three possible allocation criteria, namely effort, waiting time, and need, on a 5-point scale, ranging from 1 (not fair at all) to 5 (very fair). The criteria were presented in the same order to all participants. Each participant received either a positive [or a negative] framing of the distributive justice situation in the following manner:

Here are some possible criteria in order to decide who of the applicants should be given [deprived of] a child for adoption. Please mark your agreement regarding the fairness of the criteria on the scale from 1 (“not fair at all”) to 5 (“very fair”).

Effort: The applicants who should be given [deprived of] a child for adoption are those who did [not] make efforts in the past and tried [did not try] other means to become parents.

Waiting-time: The applicants who should be given [deprived of] a child for adoption are those who were waiting the most [least] time for adoption.

Need: The applicants who should be given [deprived of] a child for adoption are those who did not have [had] other offspring.

Procedure

The questionnaires were distributed in various locations where they could be completed in comfort. Participants randomly received either a positive or a negative version of the questionnaire. The research was presented as examining attitudes toward child adoption allocation criteria, and participants were told that their responses were anonymous, and that there were no correct or incorrect answers.

Results and Discussion

One participant did not complete the questionnaire and was removed from the analysis. The first hypothesis was that attribute framing would affect the perceived fairness of the three allocation criteria. Table 1 presents the perceived fairness of the three allocation criteria for the positive and negative framing conditions. As hypothesized, participants perceived the three criteria as more fair in the positive versus the negative framing condition. Attribute framing effect sizes (Cohen’s d) were 0.23, 0.35, and 0.58, for the respective waiting time, need, and effort allocation criteria. The differences were statistically significant with respect to the need and effort allocation criteria, but not for waiting time. The internal consistency measure of the answers to the three allocation criteria (Cronbach’s α) was .54. Thus, we averaged participants’ perceived fairness ratings across the three allocation criteria to form an average perceived fairness index.

Table 1.

Perceived Fairness of the Allocation Criteria for the Two Framing Conditions (Using a 5-Point Scale).

Framing	Effort	Waiting Time	Need	Average
Positive (n = 60)	3.65 (1.1)	3.57 (1.1)	3.65 (1.3)	3.62 (0.8)
Negative (n = 56)	2.96 (1.2)	3.32 (1.0)	3.18 (1.4)	3.15 (0.9)
Effect size (d)	0.58	0.23	0.35	0.55
t(df = 114)	3.14*	1.23	1.91*	2.97*

Note. Standard deviations are given in parentheses.

* p < .05.

The second hypothesis was that agreeableness and conscientiousness would moderate the effect of attribute framing on the perceived fairness of the allocation criteria. The averages of agreeableness and conscientiousness were 3.79 and 4.04 (SDs = 0.43, 0.50), respectively. The correlation between the two personality traits was statistically significant, r = .48, t(114) = 5.88, p < .001. The hypothesis was examined separately for each of the three criteria and for the average perceived fairness index. The pattern of the results was similar, and for simplicity of presentation the following presents only the results for the average perceived fairness index.

Figure 1 presents the average perceived fairness index (on the y-axis) as a function of agreeableness (on the x-axis of Figure 1a) and as a function of conscientiousness (on the x-axis of Figure 1b) for positive and negative framing conditions. Figure 1 further presents the regression lines predicting the average perceived fairness index from the personality trait for positive and negative framing conditions separately. Figure 1 shows similar pattern for both personality traits: The differences between the regression lines for the positive and negative conditions are negligible for low values of the personality traits, and they increase as the value of the trait increases. Stated otherwise, while those scoring low on agreeableness or conscientiousness were relatively immune to attribute framing effect, higher scoring participants were more susceptible to the effect of attribute framing.

Figure 1.

Perceived fairness of the allocation criteria for adoption as a function of (a) Agreeableness and (b) Conscientiousness for positive and negative framing conditions.

In order to examine the statistical significance of these results, framing was coded such that 0 represented negative framing and 1 represented positive framing (dummy coding). In addition, two product terms were computed for the dummy coded framing variable with agreeableness and conscientiousness (i.e., Framing × Agreeableness, Framing × Conscientiousness).

For each of the two personality traits, a hierarchical linear regression analysis predicted the average perceived fairness index using two steps: In Step 1, only framing and the personality trait were entered; and in Step 2, the product term was entered as well. Thus, in Step 1, the contribution of framing in predicting the perceived fairness is examined when the personality trait is held constant, while in Step 2, the possibility that the slopes of the positive and negative framing regression lines are different is examined.

Table 2 summarizes the results of the regression analyses. Table 2 shows that in Step 1, framing had a significant effect on the perceived fairness, β values of .27, .27, t(113) = 2.95, 2.93, p = .004, .004, controlling for agreeableness and consciousnesses, respectively. In contrast, the personality traits did not predict the perceived fairness when framing was held constant. Adding the product term to each respective personality trait at Step 2 increased the multiple correlations (from 0.27 to 0.34 and from 0.27 to 0.30). The product term Framing × Agreeableness was significant, but the product term Framing × Consciousnesses was not, β values of 2.33, 1.02, t(112)= −2.33, −1.01, p = .02, .17, respectively.

Table 2.

Summary of Hierarchical Regression Analysis for Predicting the Average Perceived Fairness Index of the Three Allocation Criteria From Framing, Personality Trait (Step 1), and Their Product Term (Step 2).

	Agreeableness			Conscientiousness
Predictors	b	SE(b)	β	b	SE(b)	β
Step 1^a
Framing	0.47*	0.16	0.27*	0.46*	0.16	0.27*
Personality trait	0.06	0.19	0.03	−0.09	0.16	−0.05
Step 2^b
Framing	−2.81*	1.42	−1.62*	−1.29	1.29	−0.74
Personality trait	−0.27	0.23	−0.13	−0.29	0.22	−0.17
Framing × Personality Trait	0.87*	0.38	1.90*	0.43	0.32	1.00

Note. N = 116.

^a R = .27 and .27 for agreeableness and conscientiousness (respectively). ^b R = .34 and .30 for agreeableness and conscientiousness (respectively).

*p < .05.

These results generally replicate previous studies in showing that attribute framing affects the perceived fairness of effort, waiting time, and need. Moreover, they suggest that this framing bias is moderated by personality traits. Attribute framing had a greater effect on the perceived fairness of the allocation criteria of participants scoring high on agreeableness or conscientiousness, compared to those scoring low, who were relatively immune to framing. However, the product term reached statistical significance only with respect to agreeableness but not for conscientiousness. In addition, Experiment 1 used one specific vignette, namely, the adoption scenario. Experiment 2 was designed to (a) replicate the results of Experiment 1; (b) generalize the hypothesized moderation to other distributive justice vignettes; and (c) examine whether agreeableness and conscientiousness moderate framing bias in nondistributive justice context.

Experiment 2

Method

Participants

Participants were 128 undergraduate students in an undergraduate program (M = 23.5; SD = 1.8; 90% were women) who participated in the study as a course requirement.

Materials

Agreeableness and Conscientiousness Questionnaire

The questionnaire was identical to the one used in Experiment 1. The internal reliabilities were similar to the values obtained in Experiment 1 (.79–.80).

Distributive Justice Questionnaire

Comprised of three distributive justice scenarios: the scenario used in Experiment 1, a scenario of allocating lung transplant to patients (adapted from Gamliel & Peer, 2010), and a scenario of allocating new computers in an organization (adapted from Gamliel & Peer, 2006). Three allocation criteria were examined in each vignette: effort (substituted by merit in the lung transplant scenario), waiting time, and need. The three scenarios and the three criteria were presented in the above order to all participants. The perceived fairness of each criterion was marked on a 7-point scale, ranging from 1, not fair at all, to 7, very fair.

Attribute framing questionnaire unrelated to distributive justice: Participants were presented with the following text manipulating positive [negative] framing: “You know a driving instructor 85% [15%] of whose students pass [fail] their driving test the first time they take it.” Participants were asked about their intention to recommend the driving instructor to a young acquaintance of theirs on a 7-point scale, ranging from 1 (I definitely won’t) to 7 (I definitely will).

Procedure and Design

The experiment was conducted in two phases. At the first phase, following an invitation sent to 160 students, 128 participants completed the agreeableness and conscientiousness questionnaire online and randomly received either a positive or a negative version of the distributive justice questionnaire. At the second phase, the same 160 students were invited to the lab to complete the attribute framing questionnaire unrelated to distributive justice. Of the 128 students who participated in the online study, 110 completed the second phase. Participants were randomly allocated to the framing conditions in the second phase, independent of their allocation in the first phase. Each student had an identifying code that served to match the data from the two phases.

Results and Discussion

The first hypothesis was that attribute framing would affect the perceived fairness of the three allocation criteria in the distributive justice scenarios and the recommendation intentions in the driving instructor scenario. Table 3 shows that in 8 of the 10 cases, the responses in the positive framing conditions were significantly higher than in the negative framing conditions (attribute framing effect sizes ranged between 0.41 and 1.13); In one case, the effect size was small (0.23) and not statistically significant, and for the need principle in the computer allocation scenario, the effect was negligible (0.04).

Table 3.

Perceived Fairness of the Allocation Criteria and Recommendation Intentions for the Two Framing Conditions (Using a 7-Point Scale).

Scenario	Framing	Effort/Merit	Waiting Time	Need
Child adoption	Positive (n = 63)	4.76 (1.8)	5.17 (1.7)	5.71 (1.5)
	Negative (n = 65)	4.31 (2.0)	4.43 (1.9)	4.52 (1.6)
	Effect size (d)	0.23	0.41	0.77
	t(df = 126)	1.33	2.33*	4.34*
Lung transplant	Positive (n = 63)	3.10 (1.5)	4.78 (1.0)	5.00 (1.0)
	Negative (n = 65)	2.40 (1.4)	3.38 (1.4)	3.95 (1.3)
	Effect size (d)	0.48	1.13	0.89
	t(df = 126)	2.69*	6.38*	5.05*
Computer allocation	Positive (n = 63)	4.68 (1.1)	4.35 (1.2)	5.49 (0.9)
	Negative (n = 65)	3.95 (1.4)	3.65 (1.3)	5.46 (0.9)
	Effect size (d)	0.58	0.55	0.04
	t(df = 126)	3.13*	3.20*	0.20
		Recommendation intentions
Driving instructor	Positive (n = 55)	5.80 (1.0)
	Negative (n = 55)	4.44 (1.5)
	Effect size (d)	1.07
	t(df =108)	5.63*

Note. Standard deviations appear in parentheses.

*p < .05.

According to the second hypothesis, agreeableness and conscientiousness were expected to moderate the effect of attribute framing on the perceived fairness of the allocation criteria, but not on the recommendation intentions in the driving instructor vignette. The averages of the agreeableness and conscientiousness personality traits were 3.77 and 3.86 (SDs = 0.54, 0.52, respectively). The correlation between the two personality traits was statistically significant, r = .41, t(126) = 6.53, p < .001. These results are similar to the ones obtained in Experiment 1.

The internal consistency measure of the answers to the nine allocation criteria (Cronbach’s α) was .66. In accordance with the analysis of Experiment 1, an average index of perceived fairness was calculated for each participant as an average of the responses to all nine questions.

Figures 2 and 3 present the average perceived fairness index (on the y axis of Figure 2a and b) and the recommendation intentions in the driving instructor vignette (on the y axis of Figure 3a and b) as a function of agreeableness (on the x axis of Figures 2a and 3a) and as a function of conscientiousness (on the x axis of Figures 2b and 3b) for the positive and negative framing conditions. Figures 2 and 3 further present the regression lines predicting the perceived fairness and the recommendation intentions from the personality trait for positive and negative framing conditions separately. With respect to the perceived fairness, Figure 2a and b shows similar findings for both personality traits replicating the results of Experiment 1: The differences between the regression lines of the positive and negative conditions are negligible for low values of the personality traits, and they increase as the value of the personality trait increases. Stated otherwise, while those scoring low on agreeableness or conscientiousness were relatively immune to attribute framing effect, higher scoring participants were more susceptible to attribute framing (this pattern was found for 16 of the 18 product terms: 3 distributive justice scenarios × 3 allocation principles × 2 personality traits; the two product terms that did not present this pattern related to the need principle in the computer allocation scenario, where no framing effect was found). In contrast, Figure 3a and b shows that the personality traits did not moderate the effect of attribute framing on the recommendation intentions as framing affected the recommendation intentions similarly for low and high values of the two personality traits.

Figure 2.

Perceived fairness of the allocation criteria as a function of (a) Agreeableness and (b) Conscientiousness for positive and negative framing conditions.

Figure 3.

Recommendation intentions as a function of (a) Agreeableness and (b) Conscientiousness for positive and negative framing conditions.

In order to examine the statistical significance of these results, the framing variable and two product terms were computed in a similar manner to the one performed in Experiment 1.

For each of the two personality traits, a hierarchical linear regression analysis predicted the average perceived fairness index and the recommendation intentions using two steps: In Step 1, only framing and the personality trait were entered; and in Step 2, the product term was entered as well. Table 4 summarizes the results of the regression analyses. Table 4 shows that in Step 1, for the two personality traits, framing had a significant effect both on perceived fairness and on recommendation intentions when the personality traits were held constant. When adding the product term in Stage 2, a different pattern was found for perceived fairness and for recommendation intentions: The multiple correlations at Step 2 were increased relative to Stage 1 only for the perceived fairness (from 0.51 to 0.56, and from 0.51 to 0.54), and not for the recommendation intentions (0.48 for both steps relating to agreeableness and 0.56 in both steps relating conscientiousness). In addition, the product terms Framing × Agreeableness and Framing × Conscientiousness were statistically significant when predicting the perceived fairness and they were not statistically significant when predicting recommendation intentions.

Table 4.

Summary of Hierarchical Regression Analysis for Predicting Average Response to the Nine Allocation Criteria and to the Recommendation Intentions in the Driving Instructor Vignette From Framing, Personality Trait (Step 1), and Their Product Term (Step 2).

Dependent		Agreeableness			Conscientiousness
Variable	Predictors	b	SE(b)	β	b	SE(b)	β
Average perceived fairness index	Step 1^a
	Framing	0.79*	0.12	0.51*	0.78*	0.12	0.50*
	Personality	0.11	0.11	0.08	0.15	0.11	0.10
	Step 2^b
	Framing	−1.78*	0.82	−1.15*	−1.05	0.89	−0.68
	Personality	−0.20	0.15	−0.14	−0.06	0.15	−0.04
	Framing × Personality	0.68*	0.21	1.68*	0.47*	0.23	1.20*
Recommend-ation intentions	Step 1^c
	Framing	1.35*	0.24	0.47*	1.38*	0.23	0.48*
	Personality	0.12	0.23	0.04	0.83*	0.23	0.29*
	Step 2^d
	Framing	1.75	1.88	0.61	2.74	1.85	0.96
	Personality	0.15	0.29	0.06	0.96*	0.29	0.34*
	Framing × Personality	−0.11	0.49	−0.14	−0.35	0.47	−0.48

^a R = .51 and .51 for agreeableness and conscientiousness (respectively) [N = 128]. ^b R = .56 and .54 for agreeableness and conscientiousness (respectively) [N = 128]. ^c R = .48 and .56 for agreeableness and conscientiousness (respectively) [N = 110].

^d R = .48 and .56 for agreeableness and conscientiousness (respectively) [N = 110]. *p < .05.

The results of Experiment 2 clearly support the hypothesis that agreeableness and conscientiousness moderate the effect of attribute framing on the perceived fairness of allocation criteria: Participants who score high on these personality traits were more susceptible to attribute framing bias, while those who score low were relatively immune to it. In contrast, the attribute framing effect found in the driving instructor scenario was not moderated by either of the personality traits.

General Discussion

The attribute framing effects found in this research replicate previous results, showing that when distributive justice situations are framed positively (who should get the good), non-egalitarian criteria such as effort, merit, waiting time, and need, are perceived as more fair than when the very same situations are presented negatively (who should be denied of the good; Gamliel & Peer, 2006, 2010). The novel contribution of the present research is the finding that this bias is moderated by personality traits that are related to social justice, namely agreeableness and conscientiousness. Participants who score high on these personality traits were more susceptible to attribute framing bias than those scoring low on these traits. These findings are consistent with previous work that showed individual differences in the susceptibility to attribute framing (Braun et al., 1997; Hardisty et al., 2010). In attempting to account for these individual differences, previous studies suggested that individuals who were more mindful of particular issues were more susceptible to attribute framing bias related to these issues, while the complementary groups were relatively immune to such bias. The current research extends this conclusion by showing that attribute framing bias is also moderated by personality traits.

Furthermore, Experiment 2 found that participants with high agreeableness or conscientiousness scores, whom we suspect are more mindful of justice issues, were more susceptible to attribute framing bias in distributive justice situations than those who scored low, but these personality traits did not moderate the framing bias in a scenario which is unrelated to social justice. These differential findings are consistent with our hypothesis that agreeableness and conscientiousness moderate attribute framing biases specifically on perceived fairness of allocation criteria. Hence, the results suggest that the mechanism underlying the moderation of framing biases may be related to the level of mindfulness or awareness of certain issues, so that lower levels result in relative immunity to biases, whereas higher levels of awareness may be associated with higher susceptibility to biases.

This conclusion is consistent with evidence from a different type of framing bias, namely, goal framing. In goal framing, messages are typically more persuasive when they present the negative consequences of not taking an action relative to the positive consequences of taking one. Several studies found higher susceptibility to goal framing biases among people who were highly involved in contexts of health behavior (Meyers-Levy & Maheswaran, 2004; Maheswaran & Meyers-Levy, 1990) and consumer behavior (Dardis & Shen, 2008; Martin & Marshall, 1999).

It should be noted that the effect of framing on highly involved people might be context dependent. If individuals do not have clear attitudes and judgments in a specific domain, they will probably be more susceptible to framing.

At the theoretical level, the current results are consistent with Levin et al.’s (1998) suggestion that positive or negative labeling of an attribute leads to information encoding that evokes corresponding favorable or unfavorable associations and consequently moderates the judgment and evaluation of relevant objects or events. People who are more mindful or aware of particular issues may be assumed to have larger and stronger networks of both positive and negative associations related to that issue. Thus, individual differences in the power of particular association networks may account for the differences in susceptibility to the framing bias.

The present findings are limited to particular scenarios and related personality traits. Future research is needed to extend the range and variability of the scenarios and personality traits tested in order to explore the variables underlying individual differences in susceptibility to other framing biases.

Finally, the findings have important practical implications as they suggest that high levels of involvement and awareness to particular issues tend to amplify the framing bias on related issues. These findings are consistent with other biases in judgment and decision making in which qualified professionals provide estimations no less biased than laymen (Peer & Salomon, 2012). Thus, just as professional knowledge and experience do not immunize experts to biases, high levels of involvement and awareness do not protect one from framing biases, but rather they seem to increase the susceptibility to such biases.

Footnotes

Acknowledgment

The authors wish to thank the associate editor, Prof. Yuichi Shoda, and two anonymous reviewers for their constructive and helpful comments that helped us to improve this article. In addition, the authors thank Shelly Ben-Horin, Sarit Berger, Lior Ginat, and Itai Ziv for their assistance in collecting data used in this research.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Braun

K. A.

Gaeth

G. J.

Levin

I. P.

(1997). Framing effects with differential impact: The role of attribute salience. Advances in Consumer Research, 24, 405–411.

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

Costa

P. T.

McCrae

R. R.

(1985). The NEO personality inventory: Manual, form S and form R. Odessa, FL: Psychological Assessment Resources.

Dardis

F. E.

Shen

(2008). The influence of evidence type and product involvement on message-framing effects in advertising. Journal of Consumer Behaviour, 7, 222–238. doi:10.1002/cb.247

Etzion

Laski

(1998). The 44-item BFF inventory in Hebrew. Israel: Tel Aviv University, Faculty of Management, the Institute of Business Research.

Gamliel

Peer

(2006). Positive versus negative framing affects justice judgments. Social Justice Research, 19, 307–322. doi:10.1007/s11211-006-0009-5

Gamliel

Peer

(2010). Attribute framing affects the perceived fairness of health care allocation principles. Judgment and Decision Making, 5, 11–20.

Goldberg

L. R.

(1990). An alternative “description of personality”: The big five factor structure. Journal of Personality and Social Psychology, 59, 1216–1229.

Hardisty

D. J.

Johnson

E. J.

Weber

E. U.

(2010). A dirty word or a dirty world? Psychological Science, 21, 86–92. doi:10.1177/0956797609355572

10.

Hirsh

J. B.

DeYoung

C. G.

Peterson

J. B.

(2010). Compassionate liberals and polite conservatives: Associations of agreeableness with political ideology and moral values. Personality and Social Psychology Bulletin, 36, 655–664. doi:10.1177/0146167210366854

11.

John

O. P.

Donahue

E. M.

Kentle

R. L.

(1991). The “big five” inventory (Versions No. 4a and 54). Berkeley, CA: Institute of Personality and Social Psychology.

12.

Kalshoven

Den Hartog

D. N.

De Hoogh

A. H. B.

(2011). Ethical leader behavior and big five factors of personality. Journal of Business Ethics, 100, 349–366. doi:10.1007/s10551-010-0685-9

13.

Levin

I. P.

Gaeth

G. J.

(1988). How consumers are affected by the framing of attribute information before and after consuming the product. Journal of Consumer Research, 15, 374–378.

14.

Levin

I. P.

Schneider

S. L.

Gaeth

G. J.

(1998). All frames are not created equal: A typology and critical analysis of framing effects. Organizational Behavior and Human Decision Processes, 76, 149–188.

15.

Maheswaran

Meyers-Levy

(1990). The influence of message framing and issue involvement. Journal of Marketing Research, 27, 361–367.

16.

Martin

Marshall

(1999). The interaction of message framing and felt involvement in the context of cell phone commercials. European Journal of Marketing, 33, 206–218.

17.

Meyers-Levy

Maheswaran

(2004). Exploring message framing outcomes when systematic, heuristic, or both types of processing occur. Journal of Consumer Psychology, 14, 159–167. doi:10.1207/s15327663jcp1401&2_18

18.

Peer

Solomon

(2012). Professionally biased: Misestimations of driving speed, journey time and time-savings among taxi drivers. Judgment and Decision Making, 7, 165–172.

19.

Piñon

Gambara

(2005). A meta-analytic review of framing effect: Risky, attribute and goal framing. Psicothema, 17, 325–331.

20.

Schmitt

D. P.

Allik

McCrae

R. R.

Benet-Martínez

(2007). The geographic distribution of big five personality traits. Journal of Cross-Cultural Psychology, 38, 173–212. doi:10.1177/0022022106297299

21.

Schmitt

Baumert

Gollwitzer

Maes

(2010). The justice sensitivity inventory: Factorial validity, location in the personality facet space, demographic pattern, and normative data. Social Justice Research, 23, 211–238. doi:10.1007/s11211-010-0115-2

22.

Shi

Lin

Wang

(2009). Linking the big five personality constructs to organizational justice. Social Behavior and Personality: An International Journal, 37, 209–222. doi:10.2224/sbp.2009.37.2.209

23.

Truxillo

D. M.

Bauer

T. N.

Campion

M. A.

Paronto

M. E.

(2006). A field study of the role of big five personality in applicant perceptions of selection fairness, self, and the hiring organization. International Journal of Selection and Assessment, 14, 269–277. doi:10.1111/j.1468-2389.2006.00351.x

24.

Tversky

Kahneman

(1981). The framing of decisions and the psychology of choice. Science, 211, 453–458. doi:10.1126/science.7455683