Does neutral proportion modulate attentional control of task conflict in the Stroop task

The Stroop interference effect is finding that it takes longer to name the colour that a word is displayed in when the word spells the name of an incongruent colour(e.g. the word RED display in blue), relative to when the word is displayed using a non-linguistic string(e.g. XXXX displayed in blue). The Stroop interference effect is commonly thought to reflect an informational conflict between the information that is supplied by the word dimension of the stimulus and the information supplied by display colour dimension of the stimulus. However, more recently, it has been proposed that the Stroop interference effect may also be due to a task conflict(Goldfarb & Henik, 2007). Task conflict refers to competition between task-sets, which are the sets of cognitive processes used to perform specific tasks. A word displayed in colour is associated with two task sets: reading and colour naming. Stroop interference occurs when the word dimension of the Stroop stimulus automatically triggers the task set of reading, which then conflicts with the task set of colour naming(Monsell, Taylor & Murphy, 2001).However, the automaticity of reading has been called into question by the finding that the Stroop interference effect can be modulated, which has been taken as evidence that the task of reading can be controlled using attention. Specifically, Goldfarb and Henik (2007) have shown that the Stroop interference effect is magnified when there is a high proportion of non-linguistic neutral trials (e.g., a row of Xs), and argued that this is because when people do not experience conflict, attentional control over the task of reading is relaxed, and they open themselves up to the interfering effects of word distractors.The present thesis examined whether the task conflict is modulated by attentional control by examining the effect of neutral trial proportion on colour-unrelated word distractors (e.g. ABBEY or CHART). Previous research by Kinoshita, De Wit and Norris ATTENTIONAL CONTROL OF TASK CONFLICT v(2017) found that colour unrelated words interfered with colour naming, but this interference was absent when the response to the colour was manual, in what would amount to a low neutral proportion condition. In the present thesis, the word interference effect was magnified when there was a high proportion of neutral trials, in both the vocal and manual Stroop tasks.Using both an analysis of RT distribution and the effect of previous trial type (the "Gratton effect"), I examined the mechanism of attentional control of task conflict.RT distribution analysis showed that the magnification of the word-interference effect in the high neutral proportion condition was constant across the RT bins in the manual task and increased in size across the RT bins of the vocal task.Previous trial type modulated the size of the word interference only in the vocal task, and only in the low neutral proportion condition.Taken together, it is suggested that increasing the proportion of neutral trials relaxes attentional control of the task set of reading and that the reactive form of control operatesonly when a high level of conflict isexperienced.