The Effects of Memory and Abstractive Integration on Children's Probability Learning

Abstract

ive integration task.-Abstractive integration was assessed by a task2 adapted partly from the Frostig Developmental Test of Visual Perception (Frostig, 1966). The task had three parts, each designed to assess an ability considered crucial in probability learning: Part A, the ability to separate or abstract one particular feature out of several different ones; Part B, the ability to focus on a particular feature and identify it despite differences in its occurrence and distracting stimuli; and Part C, the ability to identify which of a series of highly similar stimuli deviates in some minor but specified sense. Part A consisted of eight items, each depicting two or more geometrical figures superimposed on or circumscribing each other. For example, there was a square and a partially superimposed triangle, a five-pointed star intersecting with two four-pointed and two six-pointed stars, and a circle cluttered by various intersecting curves and oval shapes. The items were drawn in black ink on a white background and presented on separate pages in a small booklet (17 x 13 cm). The items were arranged by increasing level of difficulty (determined by pretest findings for number of errors, time to completion, and children's remarks). There were also eight display cards (17 x 13 cm) depicting the key figures that the child was to locate (a square, triangle, cross, crescent, six-pointed star, five-pointed star, oval, and diamond). Part B (designed to assess the ability to identify a particular form despite transformations in its size and position) included four drawings presented in random order on separate pages. Embedded in the drawings was a total of eight circles and 13 squares. These forms differed in size, location, and salience. There were also many distracting forms that resembled circles and squares, e.g., oblongs, ovals, semicircles, parallelograms, etc. Pretesting showed that children found the task to be interesting and challenging both because the key figures were cleverly hidden and because it was necessary to decide in so many problematic cases whether the form was indeed a circle or a square or merely a distracting stimulus. Part C (designed to assess the ability to identify a deviant feature) consisted of rows of five items that resembled one another. There were 10 different rows presented in random order on separate pages in a small booklet. The drawings depicted familiar