do so on three separate performance trials in three different sessions,
and then only after numerous errors in the previous trial.
With the color cues in the behavior chain removed for Subject III,
the serial position seemed to serve as an adequate discriminative
stimulus to evoke correct responses, independent of color. Subject III
did not rely on the color discrimination for chain position requiring
this subject to be sensitive to the consequences of each response
(Moerschbaecher et al. 1978).
The procedural rules were found to have no positive consequence for
either Subject I or II. For two of the subjects, the rules were
ineffective in evoking the correct response. This lack of control could
Reproduced with permission of the copyright owner. F urth er reproduction prohibited without permission.
be due to sensorimotor and perceptual impairments associated with AD
(Bartal, 1979; Burnside, 1979; Wolanin & Phillips, 1981). Subject III,
on the other hand, did benefit from procedural rule 1 (the marked chip's
serial position in each set differed) suggesting that these impairments
may be individualized.
Although the achievement of a steady state of baseline responses
was of major interest in this research, there are other behavioral
characteristics observed that should be mentioned.
The individualized reinforcers seemed to have no positive conse
quence for either Subject I or II but were effective for Subject III.
In fact, it should be mentioned that for both Subjects I and II there
was no reinforcement carry over from one session to another. An example
occurred with Subject I. During session 5, the subject met criterion
(three sequentially correct responses for the three component behavior
chain without error) on all three trials. The contingent reinforcement
was a handful of unsalted cashews. The next session, session 6, when
the subject met criterion during PT1, cashews were, again, given. The
subject brushed the cashews on to the floor stating, "There's dirt on
the table." It may be possible that reinforcement for AD victims
depends less on their past history of reinforcement and more on the
physical properties of the stimulus they perceive at any given moment.
While the reinforcing stimulus remained consistent during any given
session for Subjects I a nd II, it differed on any number of occasions
between sessions, resulting in a decreased probability of correct
responding. This, however, was not observed with Subject III. The
R eproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5 5
individualized reinforcer remained consistent for this subject through
out the duration of the research.
During preliminary training with Subject II, only two poker chips
were presented in front of the subject on the felt cloth board. One
chip was marked with the happy face and the other was blank. The
procedure started with both chips face up. After the subject pointed to
the marked chip, both chips were turned over remaining in the same
position. On one occasion the subject turned over the unmarked poker
chip 45 times in a row and finally reported to the experimenter: "You
are going to run out of chips." The subject's response did not demon
strate the perception of a relationship between the two poker chips.
Similar observations were reported by Kimble (1969) with the hippocampal
rats running to either side of the maze 50 to 75 times, and by Banks and
Russell (1967) with the difficulty of the experimental rats with at
least a 40 percent reduction in ACHE in shifting from problem to
problem.
One final observation relates to the subjects' motor responses.
Kimble (1969) observed that the hippocampal rats may be unable to
inhibit their motor responses and perhaps m ay even display a tendency
for hyperactivity. Similar effects were noted for all three subjects in
this study on different occasions. All of the subjects responded very
quickly, turning over the chips in rapid succession. Except for the
first four sessions, Subject III completed each trial in all subsequent
sessions in an average time of 1:43. In addition, all of the subjects
paced rapidly during the between-trial breaks. While out for coffee,
during the thirty minute between-trial reinforcement condition, Subject
R eproduced with permission o f the copyright owner. F urth er reproduction prohibited without permission.
was refilled, get in and out of his seat two or three times, and walk
briskly to and from the restaurant. Subject I, on the other hand, would
respond in bursts and stops similar to Kimble's findings with the
hippocampal rats. These rapid motor responses relate directly to the
large number of errors for Subject I. Conversely, Subject II responded
more slowly to the testing apparatus over time even though he would move
rather quickly during the between-trial reinforcement intervals.
For the most part, this study investigated the baseline phase of
the repeated acquisition technique. However, an intervention was
imposed for Subject III to study the effects of between-trial reinforce
ment "breaks" on positive transference. During this intervention
(sessions 7 to 19), the subject made slightly more errors in PT2 and in
PT1. While the subject did demonstrate a stable baseline performance,
these increased errors suggest that the repeated acquisition technique
is sensitive to subtle changes in behavior, over time.
Similarly, a delay of three weeks during the baseline phase for