|Titel:||Visuelle und semantische Größeninformationen in Wahrnehmung und Handlung||Sonstige Titel:||Visual and semantic magnitude information in perception and action||Sprache:||Englisch||Autor*in:||Kopiske, Karl Kunibert||Schlagwörter:||Größenverarbeitung; Zahlenverarbeitung; perception of magnitude; numerical cognition||GND-Schlagwörter:||Wahrnehmung; Visuelle Wahrnehmung; Greifen||Erscheinungsdatum:||2017||Tag der mündlichen Prüfung:||2017-06-01||Zusammenfassung:||
Visual information is vital to interact with our environment, and can give us information about our surroundings before we begin interacting with it, as well as during interaction. This visual information is often used to update what we know about an object, which may be called semantic information. Thus, it seems plausible that the two might often interact. The goal of this thesis is to give a small glimpse of how visual and semantic information of size and magnitude are processed, and when they may interact. For this thesis, several experiments were conducted to investigate whether (a) visual processing differs depending on whether the task at hand involves a direct, skilled action or conscious perception, and (b) to what degree visual features of magnitude representations matter for their processing. Study 1 examined the effect of a well-studied visual illusion of size (the Ebbinghaus illusion) on perception, as well as on grasping. Whether visual illusions, and the Ebbinghaus illusion in particular, have different effect on perception and action, respectively, has been a topic of debate for a long time, and it is a key line of evidence in the overarching debate about whether perception and action process visual size differently. To come as close as possible to settling the debate, we preregistered and conducted a large-scale experiment in four different laboratories. Controlling for potential confounds, such as context circles being perceived as obstacles, we found no difference between the Ebbinghaus illusion’s effect on action and its effect on perception. The implications of this study have been discussed in two published commentaries. I discuss these commentaries – a critique of the generalizability of our experimental design, and our rebuttal – in detail, and make the case that in a confirmatory experiment like ours, what matters is how strongly predictions are made by a theory, and how well these are tested by the experiment. Under both these criteria, our study clearly presents a very strong piece of evidence. We also conducted two studies on the ‘mental number line’, investigating two specific questions on how numeric processing in influenced by visual stimulus properties. Study 2 investigated whether Chinese representations of number would evoke a space-response association (the so-called SNARC effect) known to occur when participants respond quickly to numeric stimuli. This effects describes a phenomenon that responses tend to be faster when the location of a response is “congruent” with its location. Research suggests that this congruency effect depends largely on reading habits, such that typically, large numbers on the right and small numbers on the left display a response time advantage in European participants. The effect is thought to be generally independent of the modality of the stimuli, yet at the same time susceptible to spatial stimulus features. We also know of at least one study that did not find a horizontal SNARC effect in Chinese characters. Thus, we tested whether we would find a horizontal SNARC effect in participants from Mainland China, in Arabic digits, Chinese characters, and Chinese hand signs. We found a robust horizontal SNARC effect, corroborating the notion that the effect is independent of notation. This is commonly taken as one piece of evidence for the existence of an analog internal representation of magnitude on a so-called ‘mental number line’. Apart from being responsible for congruency effects between space and number, this mental number line has also been proposed to be compressed, such that nonsymbolic or nonverbal representations of magnitude tend to systematically underestimate differences, and larger magnitudes in general. Study 3 was designed to test whether the proposed compression of the mental number line can truly be ascribed to properties of magnitude processing, or whether it is caused by the way such representations are typically measured. To this end, we had participants complete a classic typed (verbal) magnitude estimation task, as well as a nonsymbolic magnitude estimation task that consisted of estimating the correct location of a stimulus on a ruler-like response bar. We found a robust nonlinearity and underestimation in both tasks that was not caused by task demands, and that was in fact even resistant to veridical feedback, showing that these properties should be considered a property of the processing and representation of nonsymbolic magnitude. Finally, I discuss what these results mean for our understanding of how visual and semantic magnitude information is processed by human observers.
|URL:||https://ediss.sub.uni-hamburg.de/handle/ediss/7219||URN:||urn:nbn:de:gbv:18-85405||Dokumenttyp:||Dissertation||Betreuer*in:||Franz, Volker (Prof. Dr.)|
|Enthalten in den Sammlungen:||Elektronische Dissertationen und Habilitationen|