## I

figure 3.3. Three sample objects from Carlson-Radvansky, Covey, and Lattanzi (1999). Reference objects are on the left, and the located objects (functional and nonfunctional) are on the right. On each of the reference objects, the solid line runs through the center-of-mass of the object; the dashed line runs through the object's functional part. Placements were measured as deviations from the center-of-mass line, with these deviations coded as positive if they were in the direction of the functional part. The numbers correspond to percentages of the distance for the average placements for each type of located object.

The dependent variable was the placement of the located object, measured as a deviation from a line running through the center of mass of the reference object. This deviation was coded as positive if it was in the direction of the functional part, and negative if it was in the opposite direction. If placements were based on defining the spatial terms "above" and "below" according to the geometric center of mass of the reference object, then the size of the deviations would be relatively small, with a randomly determined positive or negative sign.

The important finding was that for all objects there was an overall positive bias toward the functional part. Note that across the set of objects, the distance between the functional part and the center of mass of the object varied considerably, making it difficult to interpret the magnitude of this bias when expressed in millimeter measurements. To take this variation into account, we expressed the functional bias as a percentage of the distance of the deviation from the center-of-mass of the object, relative to the distance between the center-of-mass of the object and the center of the functional part. Within this scheme, 0% would indicate a placement above or below the center of mass, and 100% would indicate a placement directly over or under the functional part.

Figure 3.3 illustrates functional and nonfunctional located objects and lists their percentage deviations for three sample reference objects. For example, for the toothbrush, the functionally related object (toothpaste) was placed 56% of the distance away from the center of mass line toward the functional part. On average across objects, functionally related located objects were placed 72% of the distance toward the functional part. It is important to note that there was no labeling or explicit mention to the participants of the functional part of the reference object; rather, the instructions were simply to place one object above or below the other object. As such, participants could have placed the located object above or below any part of the reference object. Nevertheless, the identity of the located object and its typical function with respect to the reference object were apparently recognized by participants, and this served to bias their placements toward the relevant functional part.

On average, functionally unrelated located objects were placed at 45% of the distance between the center of mass and the functional part, exhibiting a smaller but nonetheless significant functional bias. This finding indicates that people also considered a functional interaction between these two objects. This may not be surprising, given that the nonfunctional located objects were selected to allow the same type of interaction as the functional located objects. However, these interactions were not common and would often be unsuccessful. For example, for a lamp as a reference object, the functional located object was a light bulb and the nonfunctional located object was an avocado. One could screw both the lightbulb and the avocado into the light socket, but only the former action would produce light.

For an indepth analysis of the functional bias across objects and types of interaction see Carlson and Covell (in press).

0 0