Apr11 1980 Training Fusional Convergence—Cooper & Feldman 207

100 dot matrix slides (2 X 2) photographi-

cally reproduced from those pictured in

Foundations of Cyclopean Perception.8

The RDS slides were used to create two

different pairs of stimuli. One pair con-

tained a right and left slide so that binocu-

lar viewing resulted in a dot pattern seen

with a central square in crossed disparity

(660 sec of arc). The second pair contained

two identical right eye slides paired so that

no binocular disparity was present.

The RDS slides (average luminance =

816 cd/rn2) were projected by two Kodak

model #650H Carousel projectors and were

reflected off two 3.8 x 4.4 cm moveable

front surface mirrors 8 cm from the projec-

tor. The mirrors could be automatically

turned in small increments to create a base

out fusion stimulus of from 0 to 40 prism

diopters.

After the images were reflected by the

mirrors, the images were passed through

two linear polarizing filters mounted in

front of the projector lenses at axis 45 and

135. The RDS was then rear-projected onto

a 78 cm x 68 cm piece of clear Plexiglas

sandblasted on the rear surface. Each pa-

tient was seated 40 cm from the Plexiglas

screen so that the RDS subtended an angle

of 10.2 deg to the patient.

Programming and sequencing of the

stimuli, movement of the mirrors to alter

fusional demands, delivery of reinforcers,

and recording of the responses and trials

were controlled by BRS solid state logic

(TECH SERV, INC.) and electromechani-

cal relay circuitry. Reinforcers were deliv-

ered by a BRS penny dispenser placed on

the side of the viewing screen 70 cm from

the child. Responses were made on two

BRS illuminated push panels. The panels

were 20 cm from the child and tilted 45 deg

from horizontal. The right panel contained

a centered two-dimensional darkened

square. The left panel contained no square.

Procedure

During the first experimental session, the

child wore polarizing glasses and was in-

structed to look at the screen. He was told

that he would sometimes see an RDS con-

taining an inner square “popping out”

(crossed disparity; base out) and sometimes

an RDS with no inner square (lacking ste-

reoscopic disparity). During each experi-

mental training session the two RDS’s were

presented successively. Each RDS ap-

peared an equal number of times, but in

random order, during a session. A correct

response to the stimulus (indicated by an

appropriate panel push) by the child re-

sulted in the immediate delivery of a penny

from the penny dispenser, the onset of a 3-

sec positive feedback light cue, and verbal

praise by the examiner. If that correct re-

sponse was made to the RDS containing

the inner stereo square, it also resulted in

a movement of the mirrors which increased

the convergence demand by 0.66 on the

next trial. No such mirror movement oc-

curred for a correct response to the RDS

without disparity. The first trial of each

training session always contained an RDS

in which the right and left views were su-

perimposed, i.e., an RDS with a zero con-

vergence demand from Donder’s line. An

incorrect response to either RDS resulted

in the termination of the trial and stimulus

presentation without delivery of a penny,

the 3-sec cue light, or verbal praise. In

addition, any incorrect response also re-

sulted in a decrease in the convergence

demand by 1.32 pd on the following trial (up

to the limit of zero demand). The failure of

a child to make any response within the

maximum 10-sec trial period was also con-

sidered as an error. The interval between

trials, when the screen was dark, was fixed

at 7.5 sec. Each experimental training ses-

sion lasted approximately 20 min and con-

tained about 100 trials or stimulus presen-

tations. The final level of convergence de-

mand reached was recorded (in prism diop-

ters) at the end of each session. At this time

the child was always given either a small

toy or a number of pennies. If the pennies

were selected, the exact amount was deter-

mined by the child’s level of performance

during RDS convergence training. Follow-

ing the experimental session, vectograrn

base out (BO) ranges were measured (blur,

break, and recovery points in prism diop-

ters) using the Topper vectogram (Bemell

Corp.).

RESULTS

Fig. 1 illustrates a typical initial RDS

convergence training session of a child.