206 AM J OPTOM & PHYSIOL OPTICS Vol. 57, No. 4

SILO,~ localization,b float,c motion paral lax,” etc., represents actual sensory ability. It may be the case that “correct” patient responses reflect prompts from instructional sets or clinician questions, task demands to “please” the therapist, therapist bias in interpreting patient answers, the use of monocular cues, etc. Thus, the use of line or contour stimuli does not allow the clinician to judge accurately the validity of the patient’s response.6’7

Recently, a new type of stereogram has been introduced which may solve some of the problems indicated above. These stereograms, called random dot stereograins (RDS), contain a hidden stereoscopic geometric form which can be appreciated only under conditions of bifoveal fixation.8’9 Because they contain no monocular cues, they are more likely to generate valid and reliable patient responses than traditional vectograms.

A second characteristic of traditional convergence training procedures has been their lack of concern with instituting formal operations to facilitate patient motivation. The success of therapy, especially for young patients, often hinges upon the degree of motivation they have for engaging in train.. ing exercises. Operant conditioning techniques, which emphasize the use of response contingent reinforcers to provide motivation and immediate feedback for responding, have often been demonstrated to facilitate learning and perfonnance.4,10,11 Its use in facilitating convergence training and in improving convergence ranges has not yet been explored.

A third feature of traditional convergence training involves the actual method of training. Although training usually entails the presentation of progressively increasing

 

- SILO a perceptual phenomenon noted during convergence and divergence. SI indicates that the target appears smaller and closer upon convergence, while LO indicates that the target appears larger and further away upon divergence.

b Localization: the distance perception changes which are due to vergence changes and are identified by pointing to where the target appears to be.

target appears to be suspended in space.

d Parallax: the apparent movement of the targets when the subject sways. It is due to convergence or divergence.

convergence demands and the use of informal prompts (e.g., the use of a pointer in space to attain the appropriate degree of ocular convergence), a concern for formal programmed learning is usually lacking. Operant discrimination learning, with its emphasis on differential responding and discrimination facilitation techniques, offers a specific methodology which may enhance convergence performance.4

The use of RDS in an operant conditioning paradigm incorporating discrimination facilitating techniques would seem to be an effective method for improving convergence ability and binocular vision, especially in children. The initial experiment was conducted to determine whether that procedure is more effective in improving convergence than traditional vision training or orthoptic techniques. A second study was done in order to demonstrate that convergence training, rather than just exposure to stereograms, was the variable most responsible for improvement in convergence ability.

EXPERIMENT 1

METHODS

Subjects

There were four male children between the ages of 6 and 10. Three of these patients were diagnosed as intermittent exotropes and one as an accommodative esotrope with poor fusion ranges. All had demonstrated stereopsis (660 sec of arc) using the operant RDS test described by Cooper and Feldman.4 According to the clinicians responsible for vision training, none of the children were making normal progress during training sessions in regard to improving convergence ability with vectograrns and other standard clinical training procedures. Each of the children had received at least 6 to 10 sessions of convergence training using vectograms, prior to the experimental intervention, in which fusions.! ranges were recorded (all children had experienced additional sessions where vectogram convergence performances were unreported).

Apparatus

The RDS used for training convergence during experimental sessions were 100 x