We can’t go on together with suspicious minds: Forecasting errors in evaluating the appreciation of denials

ABSTRACT

In light of public examples of false denials, it is unsurprising that people’s beliefs about denials often are negative. However, inconsistent with such beliefs, denials often are sincere, and can facilitate trust repair. To illuminate this mismatch, we advance a framework based on Construal Level Theory, to explain how people may make a forecasting error when predicting their responses to denials. In two experimental studies, we reveal that people who actually received a denial after a possible transgression (a) were less suspicious, and experienced greater trust, and (b) displayed more trusting behavior than people who imagined this. These results suggest that people underestimate the effectiveness of denials in the reconciliation process.

KEYWORDS:

• Denials
• trust repair
• forecasting error
• suspicion
• trust

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Participants’ gender and age did not differ significantly between conditions, all ps > .05.

2 Participants were willing to pay, on average, €0.40 (about $0.60) to leave ten minutes earlier, indicating they perceived laboratory minutes as a valuable resource.

3 Participants’ gender and age did not differ significantly between conditions, all ps > .05.

4 In addition to these measures of trust and suspicion, our measures also included items that directly assessed the experience of suspicion (‘I have doubts over the other’s sincerity’, ‘I am suspicious of the other’, and ‘I feel mistrust toward the other’, α = .93), and beliefs of being misled by the interaction partner (‘The other has cheated me’, ‘The other has mislead me’, ‘The other has set me up’, ‘The other has tricked me’, ‘The other has ‘stolen’ from me’, ‘The other is trying to take advantage of me’, ‘The other is trying to exploit me’, and ‘The other is trying to manipulate me’, α = .96). Analyses indicated these concepts to overlap very strongly with the reported measures of suspicion and trust (rs > .60), and revealed identical results – in terms of significant main effects, interactions and simple effects – to these measures. Thus, for reasons of parsimony, these concepts are discussed no further here; however, results for these analyses are available from the first author upon request.

5 Note that these items are not verbally identical, and thus cannot be compared directly. Nevertheless, they are closely related, and assess the same superordinate concept, namely perceived fairness. This is illustrated by the substantial correlation between these scales (r = .65, p < .001) that was observed within the no message condition – where participants receive no communication between either measurement, and these perceptions thus should be similar. Conversely, in the denial condition, where participants received a denial following the first measurement, these perceptions would be expected to deviate as a result of the message. Indeed, in this condition, a low correlation was observed between either scales (r = .16, p = .16).

6 In addition to this measure, participants were asked whether they wished to replace their partner for the subsequent round (0 = No, 1 = Yes) and whether they wanted to impose a sanctioning system that would punish unfair decisions (0 = No, 1 = Yes). For partner change, loglinear analysis indicated a significant main effect of interaction type, χ²[1] = 27.33, p < .001, but no significant interaction between interaction type and message type, χ²[1] = 1.95, p = .16. Participants were more likely to indicate wanting to replace their partner when imagining the interaction (percentage likelihood: 47.2%) than when actually experiencing it (percentage likelihood: 10.6%). For imposing a sanctioning system, loglinear analysis indicated no significant effects of our manipulations.