Model
Digital Document
Publisher
Florida Atlantic University
Description
When viewing advertisements, one could be exposed to new information about the product. During that time, one could construct ad hoc categories or simple attributes for the brand-name product. The current experiment used functional near-infrared spectroscopy (fNIRS) to measure bilateral frontal and temporal cortices to understand the contribution of constructing ad hoc categories and simple attributes on purchase intentions. The current experiment also examined the feasibility of using the tensor decomposition method compared to the grand averaging method in multidimensional fNIRS signal analysis. This is to see if tensor decomposition can maintain the pattern of hemodynamic response without losing the temporal dynamics and spatial array to find a more optimized time and regions of interest to average across. The current experiments consisted of two parts: 1) participants studied brand-name products for various ad hoc categories (Experiment 1) or various simple attributes (Experiment 2) and 2) pick for purchase brand-name products in a two-alternative forced choice purchase intention test. Three methods were used to analyze the hemodynamic response data: the grand averaging method, the tensor decomposition method, and the revised grand averaging method. The revised grand averaging method is the same as the grand averaging method but uses information from the tensor decomposition method to inform what time and channel to average across. There were behavioral priming benefits compared to products that were not studied. However, there were no differences across the study conditions. Results revealed processing benefits, not purchasing benefits, for brand-name products studied for different simple attributes as marked by changes in the left prefrontal cortex. The results from tensor decomposition revealed more details on the time and channels of interest than the grand averaging method. Findings suggest that studying different simple attributes of a brand-name product produces benefits in the purchase intention process. Also, findings suggest tensor decomposition is a feasible method for fNIRS signal analysis.
Member of