A project incorporating 15 studies on parameters of cumulative programming instructional strategies in severely handicapped populations is presented. A model was developed to allow for controlled formal investigation of potentially relevant variables including those descriptive of the subject (such as handicapping condition, race, or age), the content (areas of skill or knowledge), and the task (variables related to the actual instructional process). The model allows for design of new studies based on the results of previous ones. Fifteen studies on various aspects of subject, content, and task variables are then reviewed and data are presented in table form. (CL)