1, 2, 3 or 6 Distracter Stimuli
5 or 8 Distracter Stimuli
Time in ms
Attentional Bias for Threat Faces in Trait Anxious Children
Lauren Kelly 1, Sigrid Lipka 1,Steve Croker 2, Frances Maratos 1
1 University of Derby; 2Illinois State University
To investigate overall performance, a repeated measures ANOVA of percent correct responses was carried out with Lag (2, 3, 4, 7) and Trial Type (threat, positive, neutral) as the IVs. This revealed a significant main effect for trial type only [F(1.66, 68.14) = 23.56, p < .001, ηp2 = .37]. Pair-wise Bonferroni corrected comparisons revealed that participants performed better on threat compared with positive (p = .006) and neutral (p < .001) trials, and on positive compared with neutral trials (p = .001) (see Figure 3).
Trait Anxiety Analysis
To investigate trait anxiety, a mixed ANOVA was carried out with Lag and Trial Type as the within-subjects IVs and Trait Anxiety (high [N = 22] versus low [N = 20]) as the between-groups IV. Although there was no main effect of trait anxiety, there was a significant interaction between trait anxiety and trial type [F(1.78, 71.30) = 27.44, p = .005, ηp2 = .13]. Post hoc analyses revealed that high trait anxious participants performed better on threat compared with positive (p = .037) and neutral (p = .013) trials. Comparatively, low trait anxious participants performed better on threat compared with neutral (p < .001), but not positive (p = .25) trials. In addition, low trait anxious participants performed better on positive compared with neutral trials (p < .001) (see Figure 4).
State Anxiety Analysis
To investigate state anxiety, a similar analysis was conducted with State Anxiety (high [N = 21] versus low [N = 21]) as the between-subjects IV. There was no main effect of state anxiety and the interaction effects did not reach statistical significance.
Theories in cognitive psychology suggest that the development and maintenance of anxiety are associated with enhanced processing of threat-related information (e.g., Eysenck et al., 2007). Supporting this notion is a growing body of research indicating that highly anxious participants bias their attention towards threatening interpretations of negative stimuli, such as images of angry faces (Hughes & Kendall, 2008; for a meta analysis see Bar-Haimet al., 2007).
However, most of this research has focused on adolescents and adults, and the few studies that have been conducted in child populations are inconsistent. This suggests that there is a need to further investigate attentional biases in children. Furthermore, studies in this area typically focus on spatial biases of visual attention, despite research demonstrating that the speed of processing threat is key to the activation of defence mechanisms. This suggests that there is a need to further investigate temporal biases of visual attention.
Consequently, the aim here was to investigate the effects of high and low levels of anxiety on temporal biases of attention in children. This was done using rapid serial visual presentation (RSVP), which assesses the attentional blink (AB)effect. Based on previous adolescent/adult research, it was predicted that high anxious children would process threat-related stimuli more accurately than low anxious children .
Figure 1. The facial stimuli used displaying threat, positive and two neutral faces (N1 and N2).
- Participants and Measures
- 174 local primary school children (male=91), aged 8 to 11 years (M=9.64; SD=.92), took part in a pre-selection process:
- Trait subscale of the State-Trait Anxiety Inventory for Children (STAIC; Spielberger, 1973) administered Children’s Depression Inventory Short Form (CDI:S; Kovacs, 1992) administered Participants assigned to groups of high and low levels of trait anxiety using a tertile split Participants with high levels of depression excluded.
- Together with selection criteria, this resulted in a participant sample of 42 (23 male) children aged 8 to 11 years (M=9.57; SD=.86)
- The State subscale of the STAIC was administered during the experiment.
- Participants were assigned to groups of high and low levels of state anxiety after the experiment using a median split.
- The experiment involved an RSVP task (see Maratos et al., 2008), where participants had to determine whether threat, positive or neutral faces (see Figure 1) were embedded in a stream of distracter stimuli (i.e., pictures of scrambled faces – see Figure 2 for examples).
- All stimuli were presented consecutively at a speed of one stimulus every 133.6ms.
- Stimulus presentation was controlled using Inquisit software (www.millisecond.com).
- 120 trials in total, each containing 20 stimuli: 18 distracters and 2 targets.
- During each trial, the first target (T1) was always a neutral face and the second target (T2) was either a threat, positive, or neutral face. This resulted in three trial types:
- neutral T1 – threat T2 (threat trials)
- neutral T1 – positive T2 (positive trials)
- neutral T1 – neutral T2 (neutral trials)
- Neutral trials were counterbalanced - when T1 was N1, T2 was N2 (and vice versa).
- Lags investigated were 2 (267.2 ms), 3 (400.8 ms), 4 (534.4 ms), and 7 (935.2 ms) (see Figure 2).
- After each trial, participants were asked to indicate whether or not they had seen a face and if they had, whether it was angry, happy or neutral.
- Response accuracy was recorded automatically throughout.
- To summarise, findings demonstrate that there is an effect of emotional valence on RSVP task performance, irrespective of anxiety levels. This is consistent with both adult and child literature (e.g., de Jonget al., 2009; Visu-Petra et al., 2010 ).
- It was also found that high trait anxious children demonstrate an attentional bias for threat faces, whereas low trait anxious children demonstrate an attentional bias for emotional (i.e., threat and positive) faces per se.
- This supports adult data, which shows that anxiety is associated with enhanced processing of threat-related stimuli (for a meta analysis see Bar-Haim et al., 2007).
- Therefore, attentional biases are potentially an important factor in understanding vulnerability, maintenance and treatment factors implicated in childhood anxiety.
- Consequently, attention training used to successfully attenuate such biases in anxious adults (see Mohlman, 2004, for a review) could also benefit anxious children.
- However, it should be noted that current findings relate to non-clinical anxiety.
Figure 2. Example of a trial in which T1 was a neutral face and T2 was a threat face.
- Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M.J., & van Ijzendoorn, M.H. (2007). Threat-related attentional bias in anxious and non-anxious individuals: A meta-analytic study. Psychological Bulletin, 133, 1-24.
- de Jong, P.J., Koster E.H.W., van Wees, R., & Martens, S. (2009). Emotional facial expressions and the attentional blink: Attenuated blink for angry and happy Faces, irrespective of soical anxiety. Cognition and Emotion, 23 (8), 1640-1652.
- Eysenck, M.W., Derakshan, N., Santos, R., & Calvo, M. (2007). Anxiety and cognitive performance: Attentional control theory. Emotion, 7, 336-353.
- Hughes, A.A. & Kendall, P.C. (2008). Effect of a positive emotional state on interpretation bias for threat in children with anxiety disorders. Emotion, 8, 414-418.
- Kovacs, M. (1992). Children’s Depression Inventory Manual. Los Angeles: Western Psychological Services.
- Maratos, F.A., Mogg, K., & Bradley, B.P. (2008). Identification of angry faces in the attentional blink. Cognition & Emotion, 0, 1-13.
- Mohlman, J. (2004). Attention training as an intervention for anxiety: review and rationale. The Behaviour Therapist, 27 (2), 37-41.
- Spielberger, C.D. (1973). Manual for the State-Trait Anxiety Inventory for Children. Palo Alto, CA: Consulting Psychologists Press.
- Visu-Petra, L., Tincas, I., Cheie, L., & Benga, O. (2010). Anxiety and visual-spatial memory updating in young children: An investigation using emotional facial expressions. Cognition and Emotion, 24 (2), 223-240.
Figure 3. Mean percentage of correct responses for each trial type.
Corresponding author: L.Kelly2@derby.ac.uk
Figure 4. Effects of trait anxiety on the mean percentage of correct responses for each trial type.