Attention: Here, There and Everywhere: The Relationship between Emotional Images and Attention Distribution

Abstract

Within the field of cognitive psychology there are two opposing theoretical frameworks, the conceptual metaphor theory and the broaden-and-build theory, which attempt to explain the influence of emotionally valenced stimuli on attention distribution. The conceptual metaphor theory (Lakoff & Johnson, 1999) asserts that concrete concepts (e.g. vertical distance or brightness) are used as metaphors to scaffold mental representations of abstract concepts (e.g. love and power). These metaphors rely on sensorimotor information in order to be understood, and therefore are said to be embodied. The focus of this thesis is the “Good is Up, Bad is Down” conceptual metaphor. A central prediction of this theory is that emotionally valenced stimuli should activate the “Good is Up, Bad is Down” metaphor, and automatically shift vertical attention congruently. In contrast, the broaden-and-build theory (Fredrickson, 2004) is based on evolutionary principles (Frijda, 1986), with negative emotions associated with specific-action tendencies and positive emotions associated with diffuse-action tendencies. The main prediction derived from this theory is that when individuals are induced into positive emotions their attention is broadened, whereas when induced into negative emotions their attention is narrowed. The central aim of this thesis was to gather experimental data in support for either the conceptual metaphor theory or the broaden-and-build theory when using emotionally laden images to induce affect, compared to prior research, which has utilised valenced words.

This thesis also aimed to examine the influence, if any, of both valence and arousal of the emotional images. The literature provides conflicting views on whether these constructs are orthogonal or interconnected, and as such what effect they have on evaluative processing. To date, research examining the conceptual metaphor theory or the broaden-and-build theory has not controlled for both valence and arousal in their experimental design. Two experiments were designed to assess both aims. In Experiment 1, emotionally valenced images were presented in either the upper or lower visual field, and participants were asked to categorise the image as “positive” or “negative” by pressing a designated key on a keyboard. In Experiment 2, the emotional images were displayed in the centre of the visual field for a fixed period of time, followed by the presentation of a target letter in either the upper or lower visual field. Participants responded by pressing the corresponding key to the target letter on a keyboard. Across both experiments no shifts in attention were congruent with the “Good is Up, Bad is Down” conceptual metaphor theory, indicating that the conceptual metaphor theory is not supported when utilising images. In contrast, Experiment 2 provided experimental data in support of the broaden-and-build theory, with participants responding faster to all target letters following high valenced images regardless of their position. Finally, this thesis provides support to the notion that valence and arousal are orthogonal constructs, independently influencing higher order cognitive processes such as attention.