Anxious to detect deceit: an empirical investigation of social defense theory

ABSTRACT

Social Defense Theory (SDT) states that anxious attachment reflects an adaptive sentinel strategy, whereby anxious people should be better able to detect lies than secure people. Existing research on this issue, however, has not been able to evaluate whether heightened lie detection among anxious individuals is due to an actual ability or a bias to assume that others are lying (one that pays off when others are, in fact, lying). We addressed this issue in a study in which 254 adults had to determine whether people in videos were lying or telling the truth about their experiences. Contrary to the predictions of SDT, highly anxious people did not have a heightened ability to separate lies from truths, but were biased to assume that others were lying regardless of the authenticity of their statements.

KEYWORDS:

• Adult attachment
• lie detection
• social defense theory
• multi-level modeling
• sentinel behavior

Disclosure statement

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

Declaration of conflicting interests

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

Supplementary data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/14616734.2023.2272252

Notes

1. Deception can be defined as intentionally sending a false message or making a misleading statement to others (Ekman, 1997). Deception is fairly pervasive. In fact, there appears to be an average of one lie in every two social interactions (DePaulo & Kashy, 1998). Given such environmental uncertainty, there is value in having skills to detect deception. For instance, there are interpersonal consequences of failing to detect lies, such as harming trust and impairing individuals’ general ability to accurately detect others’ emotions (Lee, Hardin, Parmar, & Gino, 2019). It is worthwhile to note that, although lie detection may be adaptive in unpredictable environments, it is likely not the case that accurate lie detection is maladaptive in predictable, safe environment.

2. Although attachment avoidance is not the primary interest of this work, we examined it in an exploratory analysis (see Table 7). Avoidance was not related to bias or discrimination, nor was there an avoidance by deceit interaction.

3. It should be noted that the evolutionary arguments put forth by Ein-Dor et al. (2010) are not fundamental to the present goals. Past work by Ein-Dor and colleagues examines the detection of interpersonal deceit and it can be argued that social-oriented threat does not constitute “imminent” danger. With this, we recognize that the context of lies may matter for the broader theoretical idea that failing to detect deceit could have implications for one’s reproductive fitness.

4. The video clips were taken from 20 unique people. Although demographic data was not collected from these actors, we sought to ensure that the recording sample was relatively heterogeneous with respect to gender and ethnicity. The order of the stimuli was randomized (10 lies, 10 truthful statements).

5. By conceptualizing attachment along dimensions, researchers can capture variation that corresponds to meaningful differences in how people think, feel, and behave in their close relationships. The individual who scores low on both of these dimensions is secure – someone who is not anxious about rejection, who is comfortable with intimacy, and who seeks closeness to and support from relationship partners. Attachment anxiety refers to the amount of fear a person feels about rejection and abandonment within close relationships. Attachment avoidance captures the extent to which a person is uncomfortable with emotional intimacy and dependence on others.

6. Please note that state and trait anxiety were measured, but there was no intention of using it in our primary analyses.

7. The GLM in R contains a logit link function. Instead of using Y itself, we’re using the logit of Y (log[p/(1-p)]) as the response. Once we fit this model, we can then back-transform the estimated regression coefficients off of a log scale. That is, these log-odds can be converted to odds by taking the exponent of the log-odds: exp(log-odds). The log-odds can be translated into more easily interpretable model-predicted probabilities by the following transformation: exp(log-odds)/(1+exp(log-odds)).

8. There are two possibilities. Anxious people may continue to adopt a liberal threshold when the base rate of non-deceitful trials exceeds deceitful ones. Alternatively, anxious people may not show a liberal threshold, as the number of non-deceitful trials increases.

Additional information

Funding

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