Generalization of conditioned (fear) responding and the role of memory specificity.

Generalization on the basis of prior experience is a central feature of human and non-human behavior. Generalization occurs when a stimulus elicits a change in behavior due to prior experience of a regularity in the environment that did not involve that stimulus. Although adaptive in general, anomalies in generalization can give rise to a wide array of problems. For instance, elevated levels of generalization have been shown in individuals suffering from an anxiety disorder.

In the first part of this dissertation, generalization is approached from a memory perspective. After a general introduction describing how different cognitive processes such as attention and memory may influence generalization (Chapter 1), we report on a study investigating generalization in relation to working memory (Chapter 2). In a community sample, we found that the extent of generalization was a function of individual differences in the capacity to filter out irrelevant information from access to working memory. In Chapter 3, generalization is investigated in relation to autobiographical memory specificity, referring to memory for personally experienced events. In a sample of individuals suffering from an anxiety disorder, we found that individuals who had difficulty reporting specific memories of personally experienced events showed impaired discrimination learning in a predictive learning task. No differences in generalization were found. Further, we used an experimental approach to memory and manipulated the retention interval between the encoding of a learning experience and the test of generalization in a face recognition paradigm (Chapter 4), and a fear generalization paradigm (Chapter 5). We did not find convincing evidence for changes in generalization over time in either of these studies.

In the second part of this dissertation, we report on longitudinal studies that focus on the identification of premorbid markers of (risk for) psychopathology. In Chapter 6, we investigated the exact nature of the relationship between discrimination learning, generalization and anxiety symptomatology. Although both impaired discrimination learning and elevated levels of generalization have previously been shown in individuals suffering from clinical or subclinical anxiety, it remains largely unclear whether these learning abnormalities represent antecedents or consequences of pathological anxiety (vulnerability factors or diagnostic markers, respectively). In a six month longitudinal study, we found that impaired discrimination learning and elevated generalization predicted higher levels of self-reported anxiety at follow-up. The final study of this dissertation (Chapter 7) deals neither with generalization, nor with memory, but nevertheless fits nicely within our efforts to identify individuals at risk for psychopathology. More precisely, we report on a questionnaire on attentional control that we helped to develop, and discuss its capacity to predict changes in diurnal cortisol values after exposure to a prolonged psychosocial stressor. Disturbances in diurnal cortisol secretion in response to psychosocial stressors have been implicated in – and proposed as a risk factor for – several psychological disorders, including major depressive disorder, and anxiety disorders. We found that individuals who reported low attentional control in the presence of emotion at Time 1 exhibited elevated cortisol responding at Time 2, following academic examination. Finally, we end this dissertation with a general discussion of our findings, and discuss several research topics in a series of freestanding commentaries (Chapter 8).