An investigation into how emotional words affect processing in the emotional Stroop task

Masters Research - Master of Clinical Psychology (MClinPsych)Scope: The emotional Stroop effect (ESE) is calculated as the difference in reaction time between classifying the print colour of emotional (e.g., SAD) and non-emotional (PAD) words. Since participants focus on colour and ignore the emotional content, the existence of ESE demonstrates an automatic attentional bias towards emotional stimuli. Recent literature has questioned if attention bias is automatic. ESE literature thus far has relied on one particular method of analysing ESE data, which has likely influenced our understanding of automaticity in the ESE. The ESE is typically calculated by subtracting the mean response time of the emotional condition from the mean response time of the neutral condition. A common, implicit assumption is made that participants process all words despite their detrimental effect on performance and therefore processing is obligatory. However, this conclusion is based on a difference between two collections of means and does not have explanatory power to determine whether emotional processing occurs on each trial. Processing may not be obligatory on each and every trial and yet ESE can still be observed. Individuals could process emotions on some trials and successfully ignore the word’s content on other trials. EST data may in fact be a combination of these two processes that have not been partitioned out. Alternatively, processing may in fact occur on all trials with each word varying on a spectrum from shallow processing to deeper level of processing (Craik & Lockhart, 1972). Thus three options exist to explain emotional processing in the task; participants process all items, participant process some items, or participants process items depending on relevance for performance. A novel task is offered to discriminate between these options. Purpose: We tested whether processing emotional stimuli is obligatory, non-obligatory or task dependent by applying a novel, forced-processing task. In the novel forced-processing task, participants must identify both the ink colour and the emotionality of words (emotional or non-emotional). Participants are forced to read and engage with the emotional content of every item. This task is then compared with a control emotional Stroop task. The control task involves font discrimination (italic or not italic) yet does not require judgment about emotional content; participants classify the colour of print and identify if any letters are in italics. The goal of the current study is to discriminate between three alternative views of emotional processing in the emotional Stoop task; obligatory, non-obligatory or task dependent by applying a novel task. A comparison between performance in the forced-processing task and the control task may offer greater insight into mechanisms underlying emotional processing. Importantly, the three theories of interest each predict a different pattern of results. Experiment 1 tests which of the predicted patterns of interaction is supported by data. Experiment 2 then replicates Experiment 1 with minor methodological changes, which allow exploration of slow vs fast effects (Sharma & McKenna, 2004) and offers some clarification regarding the role of implicit and explicit processing. Methodology: Fifty-five participants, across two experiments, completed the control and forced-processing tasks. Each participant performed both the control emotional Stroop task and the forced-processing emotional Stroop task. On conclusion of the second task, participants completed the item classification questionnaire, the BDI-II and DASS. A 2 by 2 within-subjects analysis of variance (ANOVA) was conducted on the variables task (control, forced) and condition (emotional, non-emotional) with reaction time (RT) as the dependent measure. Results: Results were consistent across both experiments. Data revealed an inverse ESE in the forced task but not in the control demonstrating that response time for emotional items were facilitated when disengagement was not required. Results were consistent with the non-obligatory and delayed disengagement view of emotional processing with a significant ESE in the control task and a significant but reversed ESE in the forced task. General Conclusions and Implications: We concluded that emotional processing does not occur on all trials, supporting a non-obligatory view of processing, and the ESE may be driven by stimuli disengagement. Results suggests that, rather than an automatic process, participants process items differently when the task forces emotional processing. Not only was there a change in ESE between tasks, the change in instruction produced a reversed ESE in both experiments. This suggests that participants process emotional items faster when emotional processing is forced or required for performance (forced-processing task) than when emotional processing (control task) is not required. Implications of the Larger Work: The finding that some but not all emotional stimuli draw attentional bias is significant as it undermines the foundations of the ESE analysis. Two consequences are evident. Firstly, inferences made based on the ESE, both theoretical and applied, ought to now come under investigation. Secondly, changes to the emotional Stroop task are necessary to ensure consistency in item processing depth and frequency. Without methodological modifications the common ESE analysis, a comparison of means, is fundamentally flawed