The aim of this thesis was to explore the interaction of attentional and emotional effects on the modulation of the startle blink reflex at long lead intervals. Extensive research has been conducted on the effects of attention and emotion on the startle reflex, however, relatively little research has examined the combined effect of these two processes. Attentional demands were held constant in the present series of experiments while emotional valence and competing processing demands were systematically varied. Experiments 1 and 2 (Series 1) focussed on the effect of variations in the emotional valence and significance of attended foreground stimuli on attentional startle modulation. Experiments 3,4 and 5 (Series 2) focussed on the effect of variations in concurrent resource demands and emotional valence on attentional startle modulation.
A discrimination and counting task, in which participants were required to count the number of longer-than-usual presentations of one visual foreground stimulus, and ignore presentations of the other, served as the attentional manipulation, while startle was elicited at long lead intervals (3.5 or 4.5 s).
In Experiment 1(N = 72) an aversive differential conditioning procedure manipulated the negative valence associated with the foreground stimuli in the attention task. Startle facilitation appeared greater during an explicitly attended negatively valenced stimulus than during a negative stimulus alone, however, the results also suggested that attention might simply have overshadowed emotional processing or the sensitivity of startle to negative valence. Further, it was unclear to what extent the pattern of results in Experiment 1 was dependent upon the aversive nature of the conditioning procedure used. Hence, Experiment 2(N= 72) examined the modulation of startle by attention, in conjunction with a non-aversive conditioning procedure. While startle latency seemed to be sensitive to both aversive and non-aversive conditioning procedures, startle magnitude modulation did not. These results suggested that startle latency facilitation during conditioning (aversive or non-aversive) reflects the non-affective elements of conditioning, such as information processing and arousal, whereas startle magnitude seems to be sensitive only to the affective elements of conditioning. Although the results of Experiment 2 indicated that conditioning had been poorly differentiated between the CS+ and the CS- in Experiment 1, comparison of Experiments 1 and 2 demonstrated that attention facilitated startle both when the attended stimulus was associated with negative valence, and when it was not.
In Experiment 3 (N = 32) the discrimination and counting task was performed concurrently with a memory task which imposed either a light or heavy load. Startle was facilitated during attended stimuli relative to ignored, but only under light memory load conditions. It was suggested that two processes contributed to the lack of differential attentional startle modulation during heavy loads. First, startle was facilitated during discrimination and counting task stimuli by increased arousal under the heavy memory load, and second, insufficient resources were available for the task, reducing startle facilitation during the attended stimulus.
Experiment 4(N= 32) examined attentional startle modulation in an aversive or non-aversive context. Shock stimuli were presented randomly during the experiment to increase the aversiveness of the context. The results indicated no difference in attentional modulation of startle in aversive or non-aversive contexts, however, the effectiveness of the contextual valence manipulation was in question. In contrast, in Experiment 5 (N= 32), self-report and ITI startle measures indicated that negative valence was higher during threat relative to safe conditions during the threat of shock procedure used to manipulate negative valence. Startle was facilitated during attended relative to ignored stimuli, and this attention effect did not vary under threat or safe conditions, suggesting the independence of attentional and emotional effects on startle.
The results of Series 1 and Series 2 suggest that attentional startle modulation during an explicit, directed attention task is not affected by emotional valence, whether prompted by the affective qualities of the attended stimulus, or of the context. Attention and emotion appear to facilitate startle, both separately, and in combination. However, competition for cognitive processing resources with a secondary task may reduce the differential attentional modulation of startle.
It was noted that the independence of the effects of attention and emotion on the startle reflex observed in these experiments may not extend to conditions of extreme negative valence, or more pronounced cognitive processing demands. Further, this research highlighted a number of issues worthy of further investigation, including: the conditions under which non-aversive startle conditioning will emerge; dissociations between the latency and magnitude measures of the startle reflex; and the interaction between arousal and attentional effects on startle modulation. It was also suggested that future startle modulation research would benefit from the development of a theoretical framework of long-lead startle modulation able to incorporate both the modality-specific and modality-nonspecific effects of attention, in addition to the combined effects of attention and emotion on the startle reflex at long-lead intervals.