THINKING THROUGH DRAWING: PRACTICE INTO KNOWLEDGE

THINKING THROUGH DRAWING: PRACTICE INTO KNOWLEDGE 65

Justin Ostrofsky

that just because some visual information has been

selected to be processed, it does not follow that the

information that has not been selected to be the focus

of attention is not processed at all. In order to reduce

the misperception and action errors associated with

the processing of non-selected information, individ-

uals would presumably have to more strongly focus

their attention to processing the selected information

while simultaneously suppressing the processing of

non-target information.

Does the degree to which non-selected visual

information is processed actually influence draw-

ing accuracy? If the processing of non-selected

visual information has the potential to cause errors

in both perception and action, do individual dif-

ferences in the ability to suppress the processing of

non-selected, ignored information predict drawing

accuracy performance? Some preliminary results

from our recent experiment (Ostrofsky, Kozbelt, &

Seidel, in press) speak to this point. Specifically, we

measured the degree to which our artist and non-

artist participants experienced the size constancy

effect. In a size matching task (Murray, Boyaci, &

Kersten, 2006), participants saw two circles on the

computer screen (see Figure 4a); the upper circle

on the screen was always the target, and partici-

pants used arrow keys on the computer keyboard

to manipulate the size of the lower circle to match

the size of the target. Participants were explicitly

instructed to focus on matching the actual size of

the circles—that is, if they were measured on the

computer screen—rather than their interpretation

of the size of the circles.

Two conditions were tested. In the depth condi-

tion, the circles were shaded to suggest spherical

forms and were presented against a textured, con-

verging perspective background to give the illusion

that the upper, target circle was more distant than

the lower, manipulated circle. To the extent that

viewers are unable to overcome size constancy, the

manipulated circle should be made larger than the

target circle, to offset the perceptual interpretation

that a more distant object of approximately equal

retinal size should itself be physically larger. In the

non-depth condition, both circles were shown in a

uniform shade of gray matching the overall value of

the spheres in the depth condition. The background

likewise maintained the same contrast of light and

dark and included a similar texture as in the depth

condition; however, no converging perspective or

any other depth cues were included.

For both versions of the size matching task, the

performance was measured by dividing the size of

the manipulated circle by the size of the target cir-

cle. Thus, a proportion of 1 indicates a perfect size

match and a proportion greater than 1 indicates that

the participant adjusted the manipulated circle to

be larger than the actual size of target circle, which

would indicate the experience of a size constancy

effect in that participants perceived the target size to

be larger than it actually appeared.

Figure 4b shows the performance of artist and

non-artist participants in the depth and non-depth

versions of this task. As can be seen, both artist and

non-artist subjects experienced the size constancy

effect, as they made reliably larger size matching

errors of the predicted direction in the depth con-

dition compared to the non-depth condition. How-

ever, artist subjects made significantly smaller size

constancy errors in the depth condition than non-

artists. (Note that this is not a reflection of an over-

all stronger ability to match the size of two objects,

as artist and non-artist subjects did not differ in

their errors in the non-depth condition.) Further,

Figure 4. (a) Presents an example of the stimulus in the depth and non-depth conditions of the size match-
ing task. (b). Presents average error proportions for artist and non-artist participants in the depth and non-
depth conditions of the size matching task. The star above the depth condition comparison indicates a statis-
tically reliable difference between artists and non-artists.

(a)

(b)