Refine
Has Fulltext
- yes (23)
Is part of the Bibliography
- yes (23)
Document Type
- Doctoral Thesis (23)
Keywords
- Angststörung (6)
- Aufmerksamkeit (6)
- Funktionelle Kernspintomografie (4)
- Virtuelle Realität (4)
- fMRI (4)
- Kind (3)
- Kognition (3)
- Soziale Wahrnehmung (3)
- social attention (3)
- Angst (2)
This dissertation highlights various aspects of basic social attention by choosing versatile approaches to disentangle the precise mechanisms underlying the preference to focus on other human beings. The progressive examination of different social processes contrasted with aspects of previously adopted principles of general attention. Recent research investigating eye movements during free exploration revealed a clear and robust social bias, especially for the faces of depicted human beings in a naturalistic scene. However, free viewing implies a combination of mechanisms, namely automatic attention (bottom-up), goal-driven allocation (top-down), or contextual cues and inquires consideration of overt (open exploration using the eyes) as well as covert orienting (peripheral attention without eye movement). Within the scope of this dissertation, all of these aspects have been disentangled in three studies to provide a thorough investigation of different influences on social attention mechanisms.
In the first study (section 2.1), we implemented top-down manipulations targeting non-social features in a social scene to test competing resources. Interestingly, attention towards social aspects prevailed, even though this was detrimental to completing the requirements. Furthermore, the tendency of this bias was evident for overall fixation patterns, as well as fixations occurring directly after stimulus onset, suggesting sustained as well as early preferential processing of social features. Although the introduction of tasks generally changes gaze patterns, our results imply only subtle variance when stimuli are social. Concluding, this experiment indicates that attention towards social aspects remains preferential even in light of top-down demands.
The second study (section 2.2) comprised of two separate experiments, one in which we investigated reflexive covert attention and another in which we tested reflexive as well as sustained overt attention for images in which a human being was unilaterally located on either the left or right half of the scene. The first experiment consisted of a modified dot-probe paradigm, in which peripheral probes were presented either congruently on the side of the social aspect, or incongruently on the non-social side. This was based on the assumption that social features would act similar to cues in traditional spatial cueing paradigms, thereby facilitating reaction times for probes presented on the social half as opposed to the non-social half. Indeed, results reflected such congruency effect. The second experiment investigated these reflexive mechanisms by monitoring eye movements and specifying the location of saccades and fixations for short as well as long presentation times. Again, we found the majority of initial saccades to be congruently directed to the social side of the stimulus. Furthermore, we replicated findings for sustained attention processes with highest fixation densities for the head region of the displayed human being.
The third study (section 2.3), tackled the other mechanism proposed in the attention dichotomy, the bottom-up influence. Specifically, we reduced the available contextual information of a scene by using a gaze-contingent display, in which only the currently fixated regions would be visible to the viewer, while the remaining image would remain masked. Thereby, participants had to voluntarily change their gaze in order to explore the stimulus. First, results revealed a replication of a social bias in free-viewing displays. Second, the preference to select social features was also evident in gaze-contingent displays. Third, we find higher recurrent gaze patterns for social images compared to non-social ones for both viewing modalities. Taken together, these findings imply a top-down driven preference for social features largely independent of contextual information.
Importantly, for all experiments, we took saliency predictions of different computational algorithms into consideration to ensure that the observed social bias was not a result of high physical saliency within these areas. For our second experiment, we even reduced the stimulus set to those images, which yielded lower mean and peak saliency for the side of the stimulus containing the social information, while considering algorithms based on low-level features, as well as pre-trained high-level features incorporated in deep learning algorithms.
Our experiments offer new insights into single attentional mechanisms with regard to static social naturalistic scenes and enable a further understanding of basic social processing, contrasting from that of non-social attention. The replicability and consistency of our findings across experiments speaks for a robust effect, attributing social attention an exceptional role within the general attention construct, not only behaviorally, but potentially also on a neuronal level and further allowing implications for clinical populations with impaired social functioning.
Honest actions predominate human behavior. From time to time, this general preference must yield to dishonest actions, which require an effortful process of overcoming initial honest response activation. This thesis presents three experimental series to elucidate this tug-of-war between honest and dishonest response tendencies in overtly committed instances of lies, thereby joining recent efforts to move from a sheer phenomenological perspective on dishonest responding as being more difficult than honest responding to a precise description of the underlying cognitive processes. The consideration of cognitive theories, empirical evidence, and paradigms from different research fields – dishonesty, cognitive control and sensorimotor stage models of information processing – lay the groundwork for the research questions and methodological approach of this thesis.
The experiments pinpoint the underlying conflict of dishonest responding in the central, capacity-limited stage of information processing (Experiments 1 to 4), but they also demonstrate that cognitive control processes (Experiments 5 to 7) and the internalization of false alibis (Experiments 8 to 11) can reduce or even completely eliminate this conflict. The data reveals great flexibility at the cognitive basis of dishonest responding: On the one hand, dishonest responding appears to rely heavily on capacity-limited processes of response selection to overcome honest response tendencies alongside up- and downstream consequences of response activation and monitoring. On the other hand, agents have powerful tools to mitigate these effortful processes through control adaptation and false alibis. These results support and expand current theorizing of the cognitive underpinnings of dishonest responding. Furthermore, they are alerting from an applied perspective on the detection of lies, especially when considering the flexibility of even basic cognitive processes in the face of false alibis. A promising way to move forward from here would be a fine-grained discrimination of response activation, passive decay and active inhibition of honest representations in dishonest responding and the assessment of the adaptiveness of these processes.
Gazes are of central relevance for people. They are crucial for navigating the world and communicating with others. Nevertheless, research in recent years shows that many findings from experimental research on gaze behavior cannot be transferred from the laboratory to everyday behavior. For example, the frequency with which conspecifics are looked at is considerably higher in experimental contexts than what can be observed in daily behavior. In short: findings from laboratories cannot be generalized into general statements. This thesis is dedicated to this matter. The dissertation describes and documents the current state of research on social attention through a literature review, including a meta-analysis on the /gaze cueing/ paradigm and an empirical study on the robustness of gaze following behavior. In addition, virtual reality was used in one of the first studies in this research field. Virtual reality has the potential to significantly improve the transferability of experimental laboratory studies to everyday behavior. This is because the technology enables a high degree of experimental control in naturalistic research designs. As such, it has the potential to transform empirical research in the same way that the introduction of computers to psychological research did some 50 years ago. The general literature review on social attention is extended to the classic /gaze cueing/ paradigm through a systematic review of publications and a meta-analytic evaluation (Study 1). The cumulative evidence supported the findings of primary studies: Covert spatial attention is directed by faces. However, the experimental factors included do not explain the surprisingly large variance in the published results. Thus, there seem to be further, not well-understood variables influencing these social processes. Moreover, classic /gaze cueing/ studies have limited ecological validity. This is discussed as a central reason for the lack of generalisability. Ecological validity describes the correspondence between experimental factors and realistic situations. A stimulus or an experimental design can have high and low ecological validity on different dimensions and have different influences on behavior. Empirical research on gaze following behavior showed that the /gaze cueing/ effect also occurs with contextually embedded stimuli (Study 2). The contextual integration of the directional cue contrasted classical /gaze cueing/ studies, which usually show heads in isolation. The research results can thus be transferred /within/ laboratory studies to higher ecologically valid research paradigms. However, research shows that the lack of ecological validity in experimental designs significantly limits the transferability of experimental findings to complex situations /outside/ the laboratory. This seems to be particularly the case when social interactions and norms are investigated. However, ecological validity is also often limited in these studies for other factors, such as contextual embedding /of participants/, free exploration behavior (and, thus, attentional control), or multimodality. In a first study, such high ecological validity was achieved for these factors with virtual reality, which could not be achieved in the laboratory so far (Study 3). Notably, the observed fixation patterns showed differences even under /most similar/ conditions in the laboratory and natural environments. Interestingly, these were similar to findings also derived from comparisons of eye movement in the laboratory and field investigations. These findings, which previously came from hardly comparable groups, were thus confirmed by the present Study 3 (which did not have this limitation). Overall, /virtual reality/ is a new technical approach to contemporary social attention research that pushes the boundaries of previous experimental research. The traditional trade-off between ecological validity and experimental control thus becomes obsolete, and laboratory studies can closely inherit an excellent approximation of reality. Finally, the present work describes and discusses the possibilities of this technology and its practical implementation. Within this context, the extent to which this development can still guarantee a constructive classification of different laboratory tests in the future is examined.
Bereits vorangegangene Studien haben zeigen können, dass eine verstärke Generali- sierung von Furcht sowohl bei Erwachsenen, bei denen beispielsweise eine Angststörung oder eine PTSB diagnostiziert wurde, aber auch bei gesunden Kindern eine Rolle spielt. In unserer Studie untersuchten wir eine Gruppe Kinder und Jugendliche (n = 31, m = 25, w = 6; Alter = 13.35 ± 2.03), die eine Störung des Sozialverhaltens aufwiesen, auf die Konditionierbarkeit von Furcht und eine mögliche Furchtgeneralisierung. Diese Gruppe verglichen wir mit einer gesunden Kontrollgruppe (n = 29, m = 11, w = 18; Alter = 14.28 ± 2.43). Als Generalisierungsstimuli verwendeten wir ein Furchtgeneralisierungsparadigma mit zwei Frauengesichtern, die in vier Schritten aneinander angeglichen wurden. Zusätzlich führten wir mit beiden Probandengruppen ein Dot-Probe-Paradigma zur Objektivierung von Aufmerksamkeitsprozessen im Sinne eines Attentional Bias oder Attentional Avoidance mit emotionalen Gesichtern durch. Wir konnten eine erfolgreiche Furchtkonditionierung für beide Gruppen erreichen. Im Vergleich mit der gesunden Kontrollgruppe zeigte die externalisierende Probandengruppe eine verstärke Furchtgeneralisierung.
Hinsichtlich der subjektiven Valenz- und Kontingenzratings wurden die Unterschiede besonders deutlich. Eine verstärkte Generalisierungsneigung bei erhöhter Trait-Angst konnten wir nicht finden. Die externalisierende Gruppe zeigte im Vergleich mit neutralen Gesichtern bei den emotionalen Gesichtern insgesamt einen Attentional Bias. Am deutlichsten war dabei eine verstärkte Aufmerksamkeitslenkung hin zu glücklichen Gesichtern festzustellen. Für die gesunde Kontrollgruppe konnten wir keine Besonderheiten bezüglich der Aufmerksamkeitsrichtung finden. Weiterführende Studien sollten mit größ- eren Probandengruppen und nach Geschlecht und Alter gepaarten Probanden durch- geführt werden. Mit externalisierenden Probanden sollte ein Furchtgeneralisierungs- paradigma mit neutralen Stimuli (z.B. Ringe) gewählt werden, um eine subjektive Wertung emotionaler Gesichter bei den Ratings als Störfaktor auszuschließen. Für externalisierende Probanden sollte außerdem die Ausprägung von CU-Traits erfasst und die Dauer der Testung verkürzt oder auf zwei Termine aufgeteilt werden, um eine ausreichende Konzentrationsfähigkeit zu ermöglichen.
Humans in our environment are of special importance to us. Even if our minds are
fixated on tasks unrelated to their presence, our attention will likely be drawn
towards other people’s appearances and their actions.
While we might remain unaware of this attentional bias at times, various studies have demonstrated the preferred visual scanning of other humans by recording eye movements in laboratory settings. The present thesis aims to investigate the circumstances under and the mechanisms by which this so-called social attention operates.
The first study demonstrates that social features in complex naturalistic scenes are prioritized in an automatic fashion. After 200 milliseconds of stimulus presentation, which is too brief for top-down processing to intervene, participants targeted image areas depicting humans significantly more often than would be expected from a chance distribution of saccades. Additionally, saccades towards these areas occurred earlier in time than saccades towards non-social image regions. In the second study, we show that human features receive most fixations even when bottom-up information is restricted; that is, even when only the fixated region was visible and the remaining parts of the image masked, participants still fixated on social image regions longer than on regions without social cues. The third study compares the influence of real and artificial faces on gaze patterns during the observation of dynamic naturalistic videos. Here we find that artificial faces, belonging to humanlike statues or machines, significantly predicted gaze allocation but to a lesser extent than real faces. In the fourth study, we employed functional magnetic resonance imaging to investigate the neural correlates of reflexive social attention. Analyses of the evoked blood-oxygenation level dependent responses pointed to an involvement of striate and extrastriate visual cortices in the encoding of social feature space.
Collectively, these studies help to elucidate under which circumstances social
features are prioritized in a laboratory setting and how this prioritization might be achieved on a neuronal level. The final experimental chapter addresses the question whether these laboratory findings can be generalized to the real world. In this study, participants were introduced to a waiting room scenario in which they interacted with a confederate. Eye movement analyses revealed that gaze behavior heavily depended on the social context and were influenced by whether an interaction is currently desired. We further did not find any evidence for altered gaze behavior in socially anxious participants. Alleged gaze avoidance or hypervigilance in social
anxiety might thus represent a laboratory phenomenon that occurs only under very specific real-life conditions. Altogether the experiments described in the present
thesis thus refine our understanding of social attention and simultaneously
challenge the inferences we can draw from laboratory research.
In vorgegangenen Studien wurde bei erwachsenen Patienten mit Angststörungen eine verstärkte Furchtgeneralisierung, eine eingeschränkte Fähigkeit zur Reizdiskrimination sowie eine veränderte Aufmerksamkeitsverteilung nachgewiesen. In einer gesunden Studienpopulation konnte bei Kindern eine stärkere Furchtgeneralisierung nachgewiesen werden als bei Erwachsenen. Ihre Generalisierungsgradienten gleichen denen von Erwachsenen mit Angststörung. Möglicherweise haben gestörte Lernprozesse in der Kindheit somit langfristige Effekte auf die Entwicklung von Angststörungen. Obwohl die Vorgänge des Furchtlernens im Kindesalter entscheidend für das Verständnis von Angststörungen sind, gibt es kaum Studien in dieser Altersgruppe. Die vorliegende Studie untersucht die Zusammenhänge von Furchtgeneralisierung und Aufmerksamkeitsprozessen in einer klinischen Population mit internalisierender Störung im Kindes- und Jugendalter.
Hierzu durchliefen Kinder und Jugendliche mit internalisierender Störung (n= 49) sowie gesunde Kontrollen (n=48) im Alter von 9 bis 17 Jahre ein Furcht-generalisierungsparadigma mit Diskriminationstraining sowie einen modifizierten Dotprobe mit integriertem Eyetracking. Die Ängstlichkeit wurde mittels verschiedener Angstfragebögen gemessen. Im Generalisierungsparadigma wurden zwei weibliche Gesichter mit neutralem Gesichtsausdruck als Stimuli verwendet, die entweder mit (CS+) oder ohne (CS-) einem 95dB lauten Schrei sowie einem angsterfüllten Gesichtsausdruck gezeigt wurden. Zur Messung der Furchtreaktion wurden subjektive Ratings für Arousal, Valenz und Kontingenz erfasst, zudem wurde die Hautleitfähigkeit gemessen. Zur Auswertung des Dotprobes wurden die Reaktionszeiten und die Initialsakkade erfasst. Die statistische Analyse des Furchtgeneralisierungsparadigmas sowie des Dotprobe-Paradigmas wurde mittels Multivarianzanalysen mit Messwiederholung durchgeführt, gefolgt von t-Tests zur weiterführenden Analyse. Desweiteren wurden die Aufmerksamkeitsreaktionen von nicht-ängstlichen und ängstlichen Teilnehmern in Kategorien eingeteilt und mittels Chi-Quadrat Analysen verglichen. Zur Analyse des Zusammenhangs zwischen Furchtgeneralisierung und Aufmerksamkeitsprozessen erfolgte eine Regressionsanalyse mit einem GS Mittelwert als abhängiger Variable und der Ängstlichkeit und den Aufmerk-samkeitsprozessen als Prädiktoren.
Die Ergebnisse bestätigten eine solide Furchtkonditionierung anhand des „Screaming Lady“-Paradigmas in einer klinischen Population, dies war erkennbar an höheren Ratings für den aversiven Stimulus im Vergleich zum sicheren Stimulus in beiden Gruppen. Grundsätzlich höhere Furchtratings sowie höhere Ratings der Generalisierungsstimuli im Vergleich zum sicheren Stimulus wiesen auf eine stärkere Generalisierung in der Untergruppe mit höherem Angst-Trait innerhalb der internalisierenden Probandengruppe hin. Die Analyse der Dotprobe Daten ergab schnellere Reaktionszeiten sowie häufigere Initialsakkaden gegenüber furchteinflößenden Stimuli bei Patienten mit internalisierender Störung. Des Weiteren zeigten sehr ängstliche Probanden häufiger einen Attentional bias im Chi Quadrat Test als nicht-ängstliche Probanden. Dies wies daraufhin, dass sowohl bei Patienten mit internalisierender Störung als auch bei sehr ängstlichen Probanden ein Attentional bias gegenüber furchtrelevanten Stimuli vorliegt. Vor allem bei Kindern mit internalisierender Störung sagten die Ängstlichkeit und veränderte Aufmerksamkeitsprozesse die Ausprägung der Furchtgeneralisierung voraus. Somit kann ein Zusammenhang von veränderten Aufmerksamkeitsprozessen und Furchtgeneralisierung vermutet werden.
Due to the global aging society and the enormous global incidence and prevalence rates that will result in the coming years, Alzheimer's Dementia (AD) represents a growing challenge for the health care system. The pathogenesis, which is unclear in parts, the chronic progression of AD, which often lasts for years, as well as insufficient diagnostic and therapeutic options complicate an adequate psychotherapeutic and medical approach to the disease. To date, AD is also considered an incurable disease.
Therefore, it is essential to gain deeper insights into the early detection or even prevention of AD. Consideration of prodromal syndromes such as Mild Cognitive Impairment (MCI) can provide significant evidence about high-risk groups for AD progression and differentiate cognitively "normal" aging individuals from those with pathological cognitive decline. Thus, for example, functional Near-Infrared Spectroscopy (fNIRS) imaging helps identify early neurodegenerative processes. In contrast, potential risk factors and predictors of later-onset clinical symptoms of MCI and AD can most often be revealed and quantified via the use of neuropsychiatric test batteries.
The present thesis consists of four studies and aimed to assess and describe the pathological cognitive decline in a sample of elderly study participants (age: ≥ 70 years; N = 604 at baseline) of the longitudinal, observational, and prospective "Vogel Study" from Würzburg, Germany, who were primarily healthy at baseline, over two measurement time points approximately 3 years apart, to differentiate between healthy and diseased study participants and to define predictors of MCI/AD and longitudinal study dropout.
Studies 1 and 2 differentiated healthy study participants from MCI patients based on the baseline hemodynamic response of the parietal cortex recorded by fNIRS during the processing of a paradigm (here: Angle Discrimination Task [ADT]) for visual-spatial processing performance. Neuronal hypoactivity was found in the MCI patients, with both healthy study participants and MCI patients showing higher superior and right hemispheric activation. MCI patients had more difficulty resolving the paradigm. Thus, no evidence of possible compensatory mechanisms was uncovered in the MCI patients.
Study 3 first defined the four latent factors declarative memory, working memory, attention, and visual-spatial processing based on structural equation model (SEM) calculations of the sample using adequate measurement (in-)variant confirmatory factor models from the baseline assessment to the first of a total of two follow-up assessments after approximately 3 years. This allowed a dimensional assessment of pathological cognitive decline versus classificatory-categorical assignment (healthy/diseased) of the sample. In addition, the superiority of the latent factor approach over a composite approach was demonstrated. Next, using a mixed-model approach, predictive analyses were calculated for the prediction of latent factors at first follow-up by baseline risk factors. The sex of study participants proved to be the best predictor of cognitive change in all the cognitive domains, with females performing better than men in the memory domains. Specifically, for declarative memory, older age predicted lower performance regardless of sex. Additional predictive evidence emerged for low serum levels of Brain-Derived Neurotrophic Factor (BDNF) on lower attention performance and higher depression symptoms on lower visual-spatial processing performance.
Study 4 further reported baseline predictors of study dropout at first follow-up. Cognitive performance, as defined in Study 3 using the four latent cognitive factors, was a predictor of study dropout for cognitive decline in the domains of declarative memory, attention, and visual-spatial processing. Conspicuous dementia screening on the Mini-Mental Status Examination (MMSE) also predicted dropout.
Overall, both the use of fNIRS imaging to detect visual-spatial processing performance in the parietal cortex during applying ADT and the dimensional perspective of the neuropsychiatric test battery in the context of prediction and dropout analyses were found to be suitable for early detection research of MCI and AD. Finally, the results will be interpreted in the overall context and implications, limitations, and perspectives will be discussed.
Even though exposure-based cognitive behavioral therapy (CBT) constitutes a first-line treatment for anxiety disorders, a substantial proportion of patients does not respond in a clinically significant manner. The identification of pre-treatment patient characteristics that are associated with treatment outcome might aid in improving response rates. Therefore, the present doctoral thesis aimed at investigating moderators of treatment outcome in anxiety disorders: first, we investigated the neural correlates of comorbidity among primary panic disorder/agoraphobia (PD/AG) and secondary social anxiety disorder (SAD) moderating treatment outcome towards exposure-based CBT. Second, pre-treatment functional resting-state connectivity signatures of treatment response in specific phobia were studied. Within the first study, we compared PD/AG patients with or without secondary SAD regarding their clinical and neurofunctional outcome towards a manualized CBT treatment focusing on PD/AG symptoms. Prior to treatment, PD/AG+SAD compared to PD/AG-SAD patients exhibited a specific neural signature within the temporal lobe, which was attenuated to the level of PD/AG-SAD patients afterwards. CBT was equally effective in both groups. Thus, comorbidity among those two anxiety disorders did not alter treatment outcome substantially. This might be due to the high overlap of shared pathophysiological features within both disorders. In the second study, we assessed pre-treatment functional resting-state connectivity within a sample of spider phobic patients that were treated with massed in virtuo exposure. We found responders already prior to treatment to be characterized by stronger inhibitory frontolimbic connectivity as well as heightened connectivity between the amygdala and regions related to the ventral visual stream. Furthermore, patients demonstrating high within-session extinction exhibited pronounced intrinsic prefrontal connectivity. Our results point to responders exhibiting a brain prepared for the mechanism of action of exposure. Taken together, results highlight the major impact of pre-treatment characteristics on treatment outcome. Both, PD/AG+SAD patients as well as responders within the SpiderVR study exhibited heightened activation or connectivity within the ventral visual pathway and the amygdala. Pronounced visual processing together with enhanced executive control and emotion regulation seem to constitute a fruitful soil for successful exposure. The results provide starting points for personalized treatment approaches in order to improve treatment success in the anxiety disorders. Future studies are needed to investigate the benefit of neuroscientifically informed CBT augmentation strategies such as repetitive transcranial magnetic stimulation.
Predictability of threat is one of the key modulators of neural activity in fear and anxiety-related threat processes and there is a considerable number of studies focusing on the exact contribution of centromedial amygdala and Bed nucleus of stria terminalis (BNST) in animals as well as in humans. In this research field, some studies already investigated the differential involvement of both areas during temporally predictable and unpredictable threat processes in humans. However, these studies showed several limitations e.g. small sample size, no predictable threat conditions, no separation of anticipation and confrontation processes, which should be addressed in future studies. Furthermore, evidence for group-based inter-individual differences of amygdala and BNST activity during predictable and unpredictable threat processes have not been studied extensively.
Several studies suggest a relevant role of the amygdala and BNST activity in phobic processes in patients with specific phobia, but no study so far has investigated the exact contribution of centromedial amygdala (CM) and BNST during temporally predictable and unpredictable threat processes in specific phobia.
This thesis consisted of three studies and aimed to evaluate the exact contribution of CM and BNST during temporally predictable and unpredictable threat anticipation and confrontation with the use of an optimized functional magnetic resonance imaging (fMRI) paradigm, which aimed to solve methodological limitations of recent studies. Study 1 used a large sample of healthy participants who were grouped based on NPSR1 genotype, and study 2 and study 3 used a sample of patients with spider phobia. In sum, the results of all three studies indicated, that BNST is more relevant for anticipation processes as compared to the CM. Contrary, during the confrontation phase the CM displays a greater relevance for threat confrontation processes.
In recent years, various studies have investigated the extent to which treatment success can be predicted in patients with anxiety disorders based on pre-treatment fMRI activity. Therefore, this was investigated for the first time in study 3 in patients with spider phobia during temporally predictable and unpredictable threat processes. Results indicated that independent of temporal predictability lower anterior cingulate cortex (ACC) activity during threat anticipation and engaged BNST during threat confrontation might be benefitting factors for successful therapy response in spider phobia.
One of the features that defines humans as extraordinarily social beings is their striking susceptibility to the gaze of others. The research reported in this dissertation was undertaken to advance our understanding of the role of gaze cues in low-level attentional and higher-order cognitive processes. In particular, effects of gaze were examined with regard to three aspects of human cognition: (1) social attention, (2) social interaction and (3) social understanding. Chapter 1 consists of three manuscripts that investigate the boundary conditions of attention capture by direct gaze and how gaze direction is integrated with facial context information. Manuscript 1 and 2 suggest two necessary requirements for attention capture by direct gaze: a meaningful holistic facial context and sharp foveal vision, respectively. Manuscript 3 shows approach/avoidance-congruency effects between gaze direction and emotion expression on attention. Chapter 2 of this dissertation explores the role of gaze in more naturalistic social scenarios. Manuscript 4 demonstrates that gaze behavior during a conversation shapes our perception of another person. Manuscript 5 builds on these findings by showing that these perceptions define our willingness to act in a prosocial way towards our interaction partner. Finally, chapter 3 adopts a broader perspective on social cognition research with a special focus on methodological aspects. Manuscript 6 is a review highlighting the significance of methodological aspects in social cognition research and stressing the importance of sophisticated decisions on task and stimulus materials. Manuscript 7 introduces a new instrument for the assessment of social understanding in adolescents. Initial application in a young sample group indicates that an understanding of another person’s mental states is a capacity that is still developing throughout adolescence. Both manuscripts of this final chapter include eye tracking data that suggest a relationship between gaze behavior and social understanding, a finding that further emphasizes the complex and multifaceted nature of social cognition. I conclude from the findings of this dissertation that research can benefit from adopting a broad view in terms of methodological as well as temporal aspects in order to capture human social cognition in its entirety.