Emotional Specificity of Early Startle Reflex Potentiation
This thesis investigated the emotional processing of photographic stimuli. The startle blink reflex was used to index emotional processing, in line with a model predicting that this reflex should be larger in magnitude when elicited during negative emotional processing, and smaller in magnitude during positive processing, relative to reflexes during emotionally neutral material. The studies were based on a discrepancy between phobic and non-phobic individuals for startle elicited shortly after picture onset. In Study 1, participants rated the emotional characteristics of photographic stimuli. These ratings were used to select pictures for subsequent experiments. Study 2 tested whether early emotional startle modification was specific to phobic/high-fear participants by measuring blink reflexes in an unselected sample during emotionally positive pictures, neutral pictures, and two types of threatening pictures — threatening animal pictures, similar to those used for phobic participants, and threatening human pictures. Startle was elicited with a 95 dB white noise stimulus, either 300 ms or between 2 and 5 seconds after onset; this was a between-subjects manipulation. Blink magnitude results showed early startle potentiation for human threat but not animal threat pictures. In Study 3a probe time was manipulated within-subjects, and reflexes were compared between positive, neutral, threat, and mutilation (e.g., dead bodies) pictures. Early startle potentiation was observed only for high-fear participants viewing threat pictures. Study 3b was an attempt to replicate Study 3a with a different picture set. To ensure participants viewed each picture from onset, a fixation cross was presented for 500 ms prior to picture onset. The absence of late probe time startle modification or skin conductance response (SCR) enhancement suggested that the picture set was not emotionally engaging, and hence unsuitable for assessing startle modification. Study 4 retained the fixation cross, with emotional categories similar to Study 3 being subdivided into high and low arousal examples. Startle modification was consistent across probe times; potentiation was observed for both threat subgroups, and for high-arousal mutilation stimuli. Low-arousal mutilation stimuli blinks were never different from neutral blinks. In Study 5 a picture complexity manipulation was used to investigate emotional startle modification 150 ms after picture onset. Half of the pictures in each of three emotional categories (positive, neutral, and negative) were full-colour photographs; the remainder were monotone silhouettes of the target stimulus (e.g., a banana, a spider). The negative category consisted entirely of spider pictures, and the participant sample was limited to females with some spider fear. The results offered limited evidence for startle potentiation at 150 ms by spider pictures, in low general-fear participants (data averaged over complexity conditions). SCR enhancement for spider contents was not consistent across participants, suggesting the pictures were not emotionally engaging. The conclusions were that identification of picture emotional content is possible by 300 ms (to the extent of modifying the startle reflex) when attention is directed to the picture at onset, as indicated by significant startle potentiation for some negative picture contents, and that previously observed differences in responding between high and low fear participants may represent differences in attentional engagement.
Advisor: Knight, Robert G.
Degree Name: Doctor of Philosophy
Degree Discipline: Psychology
Publisher: University of Otago
Keywords: emotion; startle reflex; psychology
Research Type: Thesis