Scott, Graham G.
Emotion word processing: evidence from electrophysiology, eye movements and decision making.
PhD thesis, University of Glasgow.
Full text available as:
A degree of confusion currently exists regarding how the emotionality of a textual stimulus influences its processing. Despite a wealth of research recently being conducted in the area, heterogeneity of stimuli used and methodologies utilized prevented general conclusion from being confidently drawn. This thesis aimed to clarify understanding of cognitive processes associated with emotional textual stimuli by employing well controlled stimuli in a range of simple but innovative paradigms. Emotion words used in this thesis were defined by their valence and arousal ratings.
The questions asked here concerned early stages of processing of emotional words, the attention capturing properties of such words, any spill-over effects which would impact the processing of neutral text presented subsequently to the emotional material, and the effect of emotional words on higher cognitive processes such as attitude formation.
The first experiment (Chapter 2) manipulated the emotionality of words (positive, negative, neutral) and their frequency (HF – high frequency, LF – low frequency) while ERPs were recorded. An emotion x frequency interaction was found, with emotional LF words responded to fastest, but only positive LF words responded to fastest. Negative HF words were also associated with a large N1 component. Chapter 3 investigated the attention-capturing properties of positive and negative words presented above and below a central fixation cross. The only significant effects appeared when a positive word was presented in the top condition, and a negative word in the bottom condition. Here saccade latencies were longer and there were a fewer number of errors made. Chapter 4 reports an eye tracking study which examined the effect of target words’ emotion (positive, negative, neutral) and their frequency (HF, LF). The pattern of results, produced in a variety of fixation time measurements such as first fixation duration and single fixation duration, was similar to those reported in Chapter 2.
The existence of any spill-over effect of emotion onto subsequently presented neutral text was examined in a number of ways. Chapter 5 describes priming with emotional primes and neutral targets but no effect of emotion was found. Chapter 6 employed the same design as Chapter 4 but presented positive, negative or neutral sentences in the middle of neutral paragraphs. It was found that the positive sentences were read fastest, but the neutral sentences following the negative sentences were read faster than those following neutral sentences.
Chapters 7 and 8 employed a version of the Velten mood-induction tool to examine the effect of mood when reading emotional text. Chapter 7 was a replication of Chapter 4 with 4 participant groups: positive, negative and neutral mood. While the neutral group showed similar results to those produced in Chapter 4, the positive group only fixated on the positive HF words faster, the negative group showed a frequency effect within each emotional word type, but within HF words positive words were viewed for less time than neutral words.
Chapter 8 had participants read 4 product reviews and then afterwards rate each of the products on a set of semantic differentials. This was a 3 (mood: positive, negative, neutral) x 2 (message type: positive negative) x 2 (word type: positive negative). There was no effect of mood but positive messages were read quicker when they contained positive words and negative messages were read quicker when they contained negative words. Participants were asked to recommend each product to individuals in either a prevention in a promotion focus. When the focus was prevention there were additive effects of message and word type, but when the focus was positive there was an interaction, with the positive message conveyed using negative words being rated highest. The same pattern also emerged in the series of semantic differentials.
Possible mechanisms to account for these findings are discussed, including many incarnations of McGinnies’s (1949) perceptual defense theory. Future studies should possibly aim to combine the current knowledge with motivational, goal-orientated models such as Higgins’s (1998) theory of regulatory focus.
Actions (login required)