The Overall Perception Of The Material On The Class Constructor

Many philosophers and cognitive psychologists believed thatcategories in everyday life have a "family resemblance structure", buta lot of researchers found subjects in category construction experimentscould not reflect this kind of structure. Glenn Regehr & Lee R. Brookssuggested it was the methods of stimulus presentation that influencedsubjects' sorting strategies: simultaneous presentation would producemore unidimension sorting, and match-to-standards would lead to morefamily resemblance (FR) sorting. Milton & Wills thought the spatialintegrated level of stimulus was the main factor influencing subjects'sorting strategies when the match-to-standards were used. Based onprevious studies, the present study suggest that, employingmatch-to-standards procedure, if the stimulus can be perceived as a whole,subjects will prefer to unidimension sorting, otherwise, they will preferto FR sorting.This research addressed that the influence of holistic perceptionof stimulus on category construction with a match-to-standards procedure.Experiment 1 investigated the subjects' sort strategies when the stimuliwere geometric shapes which were not perceived as a meaningful wholeobject. The results showed subjects produced more FR sorting despite thelevels of the dimensions spatially integrated. In experiment 2, weinvestigated the subjects' sort strategies when the stimuli were morenatural category shapes. We found if holistic prototypes were provided,subjects would prefer to unidimension sort even the stimulus werespatially separable. The aim of experiment 3 was to discover subjects'sort strategies when the clear dimension information was provided, andthe stimuli were spatially integrated geometric shapes and naturalcategory shapes. Under this situation, subjects produced more FR sortingwhen stimuli were geometric shapes, however, they produced more unidimension sorting when the stimuli were more natural category. Inexperiment 4, we developed a new paradigm, namely single prototypematch-to-standards procedure. This paradigm would lead subjects take moreattention to dimension information. We found subjects prefered to FRsorting when the stimuli either were integrated geometric shapes orintegrated natural category.In summary, our investigation produced two main findings. First,employing match-to-standards procedure, if the stimulus can be perceivedas a whole, subjects will prefer to unidimension sorting, otherwise, theywill prefer to FR sorting. Second, single prototype match-to-standardsprocedure will help subjects pay more attention to dimension information,then they would show more family resemblance sort even the stimuli areintegrated natural category.