514
GθuτEux, thinus-blanc, and vauclair
Figure 4. Overhead view of the testing environment for Experiments 6 (small-size cue), 7 (intermediate-size
cue), and 8. The correct location (noted “C”), the rotational equivalent comer (noted “R”), and the geometrically
inappropriate near (noted “N”) and far (noted “F”) comers are shown in the figure.
data were subjected to a chi-square one-sample test in which we
compared the observed distribution in the geometrically appropri-
ate and geometrically inappropriate categories to the theoretical
frequency of an equal distribution of these two categories (i.e.,
50% of chance for each one). These results always reached statis-
tical significance, Orcas, X2(l, N = 50) = 6.5, p < .001; Krill,
X2(l, N = 50) = 20.48, p < .005. A second chi-square one-sample
test computed on the data obtained in the observed geometrically
appropriate category (Comers C and R), compared with an equal
frequency of distribution of the searches to these two corners (i.e.,
50°10 of chance for each one), indicated that the number of visits to
Corner C and to Corner R were not statistically different, Orcas,
,2(l, N= 34) = 1.06; Krill, X2(l, N = 41) = 0.22; p > .05.
Table 2 (Experiment 6) indicates the results for each subject
during the 10 fast trials. The individual scores obtained for
these 10 first trials do not statistically differ from the scores
performed during the 50 trials of Experiment 6, Orcas, X 2 = 0.78;
Krill, X2 = 0.01; p > .05. These results indicate that no improve-
ment of the monkeys' performance occurred during Experiment 6.
Discussion
The monkeys were not able to disambiguate the correct comer
from the geometrical one by using the single small cue. They still
relied exclusively on the geometric properties of the environment
to reorient. Because the cue was of the same size as the corner cues
Table 7
Number of Trials (Out of 50) During Experiment 6 for the
Subjects as a Function of Their Search Location (Correct,
Rotational, Geometrically Inappropriate Corners)
Box
|
Monkey |
C |
R |
N |
F |
|
Orcas |
20 |
14 |
8 |
8 |
|
Krill |
22 |
19 |
6 |
3 |
|
Average (%) |
42 |
33 |
14 |
11 |
Note. C = correct; R = rotational; N = near misses; F = far misses.
(Experiment 5), but located on one of the small sides of the room
(the side where the blue wall had been affixed in Experiment 2), it
thus appears that the size but not the location of the cue was useful
for monkeys to reorient.
Experiment 7
The purpose of this experiment was to further examine the
conclusion of Experiment 6, that is, whether the size of the cue is
important for the monkeys. Thus, we tested the monkeys by using
an intermediate landmark size, between those used in Experi-
ments 2 and 3 (large) and in Experiment 6 (small).
Method
Subjects. Experiment 7 was run with 2 subjects because Subject Orcas
stopped searching for reward. Between Experiments 6 and 7, a delay of 1
week occurred.
Apparatus. The apparatus was the same as used in Experiment 6
except that the landmark used in that experiment was of intermediate size
(50 cm X 60 cm) between the large one used in Experiment 2 and the small
one used in Experiment 6 (see Figure 4).
Procedure. The procedure was identical to the one used in the previous
experiment except that a 10-trial control session was given at the end of
the 50 trials run per each subject. It consisted of the same 180° rotation of
the landmark and the rewarded box as in Experiment 2. The purpose of this
control session was to check that the information used by the subjects to
reorient was only the colored cue, together with the geometry of the
experimental apparatus.
Results
Table 8 presents the number of first-choice searches performed
by the 2 subjects during the 50 test trials. For each monkey, the
data were subjected to a chi-square one-sample test in which we
compared the observed distribution in the geometrically appropri-
ate and geometrically inappropriate categories to the theoretical
frequency of an equal distribution of these two categories (i.e.,
50% of chance for each one). These results always reached a
statistically significant level, Crevet, X2(1, N = 50) = 36.28; Krill,
Al , N = 50) = 42.32; p < .001. A second chi-square one-sample