Processing efficiency in anxiety: Evidence from eye-movements during visual search
Introduction
Cognitive models of anxiety have postulated that biased attentional processing of threatening information is an important characteristic of individuals with elevated risk to develop anxiety disorders (Mogg & Bradley, 1998). These attentional biases to threat are considered to be crucial cognitive risk factors associated with heightened emotional reactivity and the etiology and maintenance of anxiety disorders. There is overwhelming empirical evidence in support of the notion that anxiety is characterized by attentional biases towards threat-related material (Bar-Haim et al., 2007, Cisler and Koster, 2010). There is also emerging research to support the prediction that attentional bias is causally involved in anxiety (MacLeod, Koster, & Fox, 2009). An important challenge for current research is to better understand the nature of this attentional bias in order to more precisely elucidate the consequences of attentional bias.
An important debate in the literature has focused on the question of whether anxiety is associated with facilitated attentional orienting towards threat or by difficulties to disengage attention from threat (Fox, Russo, Bowles, & Dutton, 2001). Although attentional bias was initially conceived of as a facilitated attentional orienting to threat (Williams, Watts, MacLeod, & Mathews, 1988) this notion was challenged by research using the spatial cueing task (Posner, 1980). In the emotional adaptation of this task a peripheral threat or neutral cue is immediately followed by a target presented at the same or the opposite location of the peripheral cue. This task has been used to examine whether presenting a target at the same location of a threat cue is associated with reaction time benefits indicating faster orienting towards the location of threat). Moreover, it can be examined whether threat causes reaction time costs when attention has to be reallocated on trials where the target is presented at the opposite location of a threat cue (impaired disengagement of threat). Studies using this task have shown that anxiety is specifically associated with impaired disengagement from threat and not enhanced attentional orienting (for a review, see Fox, 2004, but see Koster et al., 2007). However, the spatial cueing task may not be suitable to capture facilitated attentional orienting (e.g., Weierich, Treat, & Hollingworth, 2008) and potential response confounds could inflate the possibility to find disengagement effects (Mogg, Holmes, Garner, & Bradley, 2008).
Theorists therefore have been inspired to develop more fine-grained models to understand the temporal unfolding of attentional bias for threat in anxiety. One particularly influential model is the attentional control theory (Derakshan and Eysenck, 2009, Eysenck et al., 2007) a major development of the earlier processing efficiency theory (Eysenck & Calvo, 1992). A cornerstone of this theory is the distinction between performance effectiveness and performance efficiency. Effectiveness refers to an individual’s competence in doing a task (usually measured by response accuracy), and efficiency refers to the amount and the manner in which processing resources are invested in doing the task (usually measured by response latency). This theory predicts that anxiety has a greater impact on performance efficiency of tasks requiring the inhibition and shifting functions of the central executive. In inhibition, attentional control prevents attentional resources from being allocated to task-irrelevant stimuli, and in shifting, attentional control is used in a positive way to allocate attentional resources to execute the task relevant to the current goal. In relation to threat processing, the theory claims that anxiety disrupts the balance between stimulus-driven (involved in bottom-up control) and the goal-directed (involved in top-down control) systems. These two systems are generally thought to interact in their functioning (Corbetta and Shulman, 2002, Pashler et al., 2001). Anxiety is believed to increase the influence of the stimulus-driven system over the goal-directed processes, reducing attentional control.
The present study investigated whether attentional bias in anxiety is a reflection of facilitated attentional orienting towards threat or attentional impairments after threat detection. The visual search task provides a useful paradigm to examine attentional capture by threat as well as attentional disengagement. In this task, participants are provided with different type of instructions. In the present study we used the “odd-one-out” instructions where individuals were instructed to see whether a display of 8 faces contained a face (target) with a different emotional expression. By manipulating the emotional expression of the target and the crowd, one can investigate speeded threat detection as well as impaired disengagement from threat through the examination of response latencies in high vs. low-anxious individuals. For example, an angry face (target) can be presented in a display of neutral faces (crowd) to investigate speeded threat detection. Conversely, a neutral face (target) can be presented among an angry crowd to examine attentional disengagement. The visual search task has demonstrated anxiety-related attentional biases both at the level of facilitated attentional engagement to threat (e.g., Öhman et al., 2001, Rinck et al., 2005) as well as impaired disengagement from threat (Byrne & Eysenck, 1995).
One important drawback of most research using this task is that response latencies are used as the only dependent variable. It has been argued that the presence of threat could facilitate behavioral responding, inflating the chance to conclude that anxious individuals show facilitated attentional engagement with threat (see Flykt, 2006). Others have argued that threat can disrupt behavioral responding which could lead to a higher probability to find disengagement effects (Mogg et al., 2008). Thus, it is possible that manual responses do not provide an accurate index of attentional processes. Moreover, during visual search there may be several shifts of attention before as well as after target detection that are not indexed by manual response latencies. Therefore, we tracked eye-movements during visual search to obtain a more comprehensive picture of threat processing at levels of both crowd and target. This methodology allowed to examine attentional operations before detection of threat (facilitated attentional engagement) as well as after detection of threat (impaired attentional disengagement).
We used emotional expressions of faces (angry, happy, and neutral) as stimuli in a visual search paradigm. We used angry expressions as there is accumulating evidence to suggest that high-anxious individuals compared with low-anxious individuals show an enhanced attentional bias towards angry facial expressions of emotion over other emotional expressions and regard them as personally relevant (see Bar-Haim et al., 2007, Fox, 2008 for reviews). By combining information from manual responding and eye-movements it is possible to examine some of the predictions from the processing efficiency and attentional control theories and to determine whether anxiety is associated with facilitated orienting towards or impaired disengagement from threat. We predicted that threat biases attention through disruption of the inhibition function and thus cognitive processing would be hampered in high-anxious individuals upon detection of threat. This would lead to the prediction that behavioral responding will be delayed when threat has captured attention and attention needs to be disengaged from threat. Therefore in order to examine how threat processing is affected by both target and crowd our method of analysis distinguished between attentional operations before and after target detection.
Section snippets
Participants
A total of 77 participants (45 female and 32 male) were recruited via advertisements posted in University of London departments. They had a mean age of 31.16 years (SD = 10.35). All participants had normal or corrected-to-normal vision and were allowed to wear their glasses or contact-lenses if required. They were paid £5 for their contribution.
Stimuli
The stimulus images were obtained from the Parke–Waters face model (Parke and Waters, 1996, Waters, 1987). This 3-D polygonal representation was deformed
Results
Mean and median trait anxiety scores for the entire sample were 41.62 (SD = 8.89) and 42, respectively. Median splits2 were used to define participants into low anxious (N = 38) and high anxious (N = 39). Mean trait anxiety for the low-anxious group was 34.34 (SD = 4.67) and for the high-anxious group it was 49.60 (SD = 6.28), F(1, 75) = 146.34, p < .001. Less than 1% of all trials were lost due to
Discussion
We examined the attentional processes associated with visual search for emotional and neutral target and crowd faces. Related to current debates on the nature of attentional bias in anxiety, we distinguished between attentional processes before and after threat detection. There were two main findings: First, that anxiety was neither associated with facilitated attention towards or with impaired disengagement from angry facial expressions prior to target detection; and second, that anxiety was
Acknowledgements
This work was partially supported by a Royal Society Joint International Grant awarded to Nazanin Derakshan and Ernst Koster. We are grateful to Miles Hansard and Leor Shoker for assistance with programming and data extraction.
References (34)
- et al.
Mechanisms underlying attentional biases towards threat: an integrative review
Clinical Psychology Review
(2010) - et al.
The emotional Stroop interference effect in anxiety: attentional bias or cognitive avoidance?
Behaviour Research and Therapy
(1994) - et al.
A cognitive-motivational analysis of anxiety
Behaviour Research and Therapy
(1998) - et al.
Effects of threat cues on attentional shifting, disengagement and response slowing in anxious individuals
Behaviour Research and Therapy
(2008) - et al.
Threat-related attentional bias in anxious and nonanxious individuals: a meta-analytic study
Psychological Bulletin
(2007) - et al.
Trait anxiety, anxious mood, and threat detection
Cognition & Emotion
(1995) - et al.
Control of goal-directed and stimulus-driven attention in the brain
Nature Reviews Neuroscience
(2002) - et al.
Anxiety, processing efficiency and cognitive performance: new developments from attentional control theory
European Psychologist
(2009) - et al.
The facial action coding system
(1978) - et al.
Anxiety and performance: the processing efficiency theory
Cognition & Emotion
(1992)
Anxiety and cognitive performance: the processing efficiency theory
Emotion
Preparedness for action: responding to the snake in the grass
American Journal of Psychology
DMDX: a Window display program with millisecond accuracy
Behavior Research Methods, Instruments, & Computers
Maintenance or capture of attention in anxiety-related biases
Facial expressions of emotion: are angry faces detected more efficiently?
Cognition & Emotion
Do threatening stimuli draw or hold visual attention in subclinical anxiety?
Journal of Experimental Psychology: General
Cited by (43)
Social anxiety is related to increased dwell time on socially threatening faces
2016, Journal of Affective DisordersRevisiting attentional processing of non-emotional cues in social anxiety: A specific impairment for the orienting network of attention
2015, Psychiatry ResearchCitation Excerpt :While the latter refers to an individual׳s competence in doing a task (i.e. response accuracy), the former refers to the amount and the manner in which processing resources are invested in doing the task (i.e. reaction times). Accordingly, Derakshan and Koster (2010) reported that anxiety was associated with reduced performance efficiency. Yet this cannot explain why no group difference was observed at the accuracy level in the previous studies that administrated the ANT to patients suffering from disorders that are not directly related to anxiety, such as tinnitus (Heeren et al., 2014) or Wilson׳s disease (Han et al., 2014).
Exploring the function of selective attention and hypervigilance for threat in anxiety
2014, Clinical Psychology ReviewCitation Excerpt :Findings of rapid engagement with threat have typically been observed in free-viewing tasks or RT tasks with concurrent eye movement measures, which include simple visual displays containing two or four items (e.g., Garner et al., 2006; Mogg et al., 2000, 2007; Stevens et al., 2011). These findings have not typically been replicated in studies using more complex visual displays (e.g., Derakshan & Koster, 2010; Huijding et al., 2011). Furthermore, some eye tracking evidence suggests that there is capture of overt attention by task-irrelevant threat in anxiety (Miltner et al., 2004); however, more recent studies have questioned this finding by using traditional eye tracking paradigms to show that anxious individuals are able to suppress exogenous saccades to threat stimuli (Derakshan et al., 2009; Richards et al., 2012).
Relationships between attention to emotion and anxiety among a community sample of adolescents
2022, Psychological Medicine