Differential factors related to the cause and duration of attentional bias in the emotional Stroop task

The emotional Stroop task (EST) is a widely used method in demonstrating how emotional material disrupts performance on a simple task through the biasing of attention. The finding that participants take longer to identify the colour of emotional material relative to neutral material is known as the emotional Stroop effect (ESE). The ESE was thought to be relatively fast, occurring on a single trial. However, recent research has suggested that emotional, primarily negative, material may disrupt colour-naming ability beyond the time of its presentation, disrupting performance in subsequent trials. That is, the ESE may be comprised of both fast and slow components. Currently there is no consensus as to the duration of the slow disruption or the mechanisms underlying this form of biased attention. Recently, Wyble, Sharma, and Bowman (2008) developed a computational model accounting for the occurrence of both fast and slow components of biased attention within the EST. The model makes specific predictions regarding the relationship between anxiety and the occurrence of fast and slow effects within the task; however these predictions have not been tested. The purpose of the research was to test these predictions in individuals with varying levels of anxiety. In total, five experiments were conducted; two to select stimuli for use in the ESTs (Experiments 1 and 3; n = 250) and three ESTs (Experiments 2, 3, and 5; n = 317) whereby fast and slow effects were examined in individuals who were low anxious, state anxious, trait anxious, and state-trait anxious. Duration of slow effects was examined by manipulating the length of the inter-trial intervals (ITIs) between words in the EST, in addition to tracking reaction times (RTs) over a series of five positions. Furthermore, positive emotion and negative emotion words with comparable arousal ratings were utilised in conjunction with neutral words to determine whether arousal, in addition to valence, played a role in biasing attention. Results showed mixed support for the hypotheses. The first emotional Stroop experiment (Experiment 2) found no evidence of fast effects. Slow effects were noted at the 32 ms block, albeit in a position later than expected. Slow effects emerged for all individuals on Position 4 between neutral words in the positive emotion sequence versus neutral words in the negative emotion sequence. There were also unexpected patterns of responding to pure sequences of neutral words. Due to the unexpected results, a decision was made to run an additional emotional Stroop experiment (Experiment 3) utilising different neutral stimuli. This experiment found no evidence of fast or slow effects occurring. Based on the inconsistencies in results between Experiment 2 and 3, Experiment 5 was conducted utilising a contingency-free methodology that allowed for the independent assessment of fast and slow effects. Results from Experiment 5 did not find evidence of a fast effect, however, in the anxious group, participants responded to neutral words that were presented after positive emotion words significantly faster than neutral words presented after negative emotion words. Collectively, the results did not find support for a fast component of attentional bias, which was contrary to predictions. Additionally, mixed support emerged for the presence of slow effects. While disruptions in colour-identification were noted on neutral words that followed emotion words, these were generally not in the position expected and did not always implicate the expected word type. The findings of the current study do not support the predictions of the Wyble et al. (2008) model. Implications of the current findings in addition to future directions are discussed.