SECOND EXPERIMENTS

< Abstract
< First experiments
> General Conclusions

The second series of experiments focuses on the recognition of everyday common objects that are rotated in the picture plane to various degrees. Again, much research has been done on this. It took off in the early 70s when two researchers (Shepherd and Metzler, 1971) found that the time to decide whether two objects were the same or mirror reflections of each other increases linearly as one object is rotated away from the other in either the picture plane (imagine a picture rotating) or in depth (imagine walking around an object). In two series of experiments I sought to explore the nature of this phenomenon.

In the first experiments, participants are presented with 120 objects. The objects can be presented at any rotation in 60 degree steps (in clockwise rotation; 30, 90, 150, 210, 270 degrees). Each object is presented 4 times. Each time the object is presented it is presented at a different orientation and never at either 0 degrees (i.e., upright) or 180 degrees (i.e., upside down). Participants are instructed to either indicate which way the object needs to be rotated (clockwise or counter clockwise) in order to get it back to its upright position (i.e., 0 degrees) or to indicate which way the object has been rotated from 0 degrees to get to where it is. In both cases participants are to rotate the object the shorter distance. The results show that the time to respond increases as the orientation of the object increases away from 0 degrees, but that there is little or no difference in the tasks. So, it seems deciding which way something has been rotated is not more difficult than deciding which way something needs to be rotated to take it back to its normal position. We’re not sure why this is, but it could be that the process(es) involved in making either decision is the same.

The second group of experiments sought to investigate the results of the first by looking at whether the orientation of an object can be determined independently of the object and its orientation. The first experiment consists of two tasks. In one task, half of the objects are presented upright and half are presented at any other angle in 30 degree clockwise steps (e.g., 90 degrees is 3 o’clock and 270 is 9 o’clock). Participants are instructed to press ‘1’ if the object is upright and ‘2’ if it is at any other angle. In the other task, half of the objects are presented upside down (180 degrees) and half at any other angle (including upright). Here participants are instructed to press ‘2’ if it is at 180 degrees and ‘1’ if it is at any other angle. The results of these experiments show that the closer an object is to the critical angle (i.e., 0 in the first task and 180 in the second) the slower responses become. Interestingly, upright objects are confused with upside down objects and vice versa. It seems then that participants are essentially confusing the angle of the objects. That is, it appears that the angle of the objects can be determined rapidly, and that the increasing response times found are due to an increasing confusion between the angle of the objects with one response and the angle of the objects with the other response.

A further series of experiments look at this issue in more detail. This time the two separate tasks discussed above are combined into one task. Now participants are presented with objects that could occur at any orientation equally often. Participants are instructed to make one response if the object is more upright and another if it is more upside down. So here, if an object is presented at 90 degrees away from upright (i.e., 90 and 270) then it is neither more upright nor more upside down. It turns out the time to make a response increases as the angle of the object approaches 90 degrees. This reflects the idea above that as an object gets closer to the “criterion” angle the more difficult it will be to decide which response is correct. In turn this reflects the idea that it is not the angle that is hard to determine. Instead, response times increase as a result of increasing confusion between the angle of an object and the criterion angle (i.e., 90 degrees in the current experiment and 0 and 180 in the above experiment).

In summary then, it seems no extra time is needed to determine the angle of an object as it is rotated away from upright. Therefore, it seems that the results of the first experiments described here cannot be explained by an increase in time to locate the angle of the objects. So, in summary the results of that experiment are yet to be adequately understood and more will hopefully be revealed as I explore the data in more detail.

Reference: Shepard, R. N., & Metzler, J. (1971). Mental rotation of three-dimensional objects. Science, 171, 701-703.