Evolutionary Evidence: Anxiety, Attentional Biases, and the Dot Probe Detection Paradigm

Anxiety is viewed as an evolutionary system biased toward false positives of threat. This means that anxiety might be a form of fear that occurs when an eliciting event is not present or when a harmless eliciting event is present. Thus, anxiety would be adaptive in the sense that it is better to experience unwarranted fear than to NOT experience fear in the face of a dangerous stimulus. Evidence that anxiety is a prolonged fear reaction stems from work on generalized anxiety disorder and trait anxiety. Generalized anxiety disorder (GAD) is a DSM-5 diagnosis. People scoring high on trait anxiety experience stable, high levels of anxiety across situations and over time. Studies on GAD and trait anxiety participants have found that anxiety causes attentional biases. An attentional bias occurs when people maintain focus on threatening information for a longer than expected period of time or when people experience difficulty moving their attention away from threatening stimuli.

The dot probe detection paradigm is one methodology used to evaluate attentional biases. In the dot probe paradigm, participants are shown pairs of photos or pairs of words. Usually, one of the photos is a threatening face or word and the other photo is a neutral/positive face or word. A plus sign is placed in the middle of the screen and separates the two photos. During each trial, participants are instructed to focus on the plus sign. During each trial, the two photos disappear, and a small dot will remain in the area of one of the photos. Participants are instructed to press one of two buttons to indicate whether the dot replaced the photo on the left or right of the screen. Response latencies are measured to see how quickly participants hit the button once the dot appears. Faster response latencies for threatening versus neutral photos would mean the participants are already looking at the threatening photo when the dot appears. Faster response latencies for neutral versus threatening photos would mean the participants are already looking at the neutral photo when the dot appears.

Compared to control participants, trait anxious participants displayed faster response latencies to threatening faces and slower response latencies to happy faces (Bradley et al., 1997). These findings mean that anxiety is associated with vigilance for threatening stimuli, meaning it takes anxious participants a longer time to disengage from the threatening face. Further work has found that anxious participants responded faster to threatening versus neutral faces and slower to happy versus neutral faces (Bradley et al., 1998). These findings suggest that anxious participants show a bias toward maintaining attention on threats (called hypervigilance) and a bias toward avoiding happy or positive-eliciting stimuli. Conversely, non-anxious participants show a bias toward maintaining attention on happy versus neutral faces and a bias toward avoiding threatening stimuli and focusing on neutral faces. These findings have been replicated with threatening and non-threatening words (MacLeod et al., 1986).

Eye-tracking studies provide evidence of attentional biases. GAD participants more frequently and rapidly looked at the threatening stimulus first rather than the neutral/positive stimuli (Mogg et al., 2000). When the stimuli are masked happy and angry faces, anxious individuals move their eyes to the masked angry faces, indicating these attentional biases occur at the nonconscious level. Thus, fear is associated with moving attention to threat in environment, whereas anxiety is associated with difficulty disengaging from future threats in the environment.