RESEARCH IS REPORTED HERE CONCERNING THE ASSUMPTION THAT ITEMS IN A LIST MUTUALLY AFFECT EACH OTHER IN VERBAL LIST-LEARNING. THE AUTHOR CONSIDERS BOTH THE MODE OF THE DATA ANALYSIS AND THE METHOD OF S-R (STIMULUS RESPONSE) PRESENTATION FOR A NUMBER OF RESTRICTED THEORETICAL ASSUMPTIONS INVOLVING ITEM INTERACTIONS IN S-R LIST-LEARNING EXPERIMENTS. IN SUCCEEDING CHAPTERS--(1) AN ANALYSIS IS MADE OF HOW BEST TO USE DATA IN LIST-LEARNING EXPERIMENTS SO AS TO BEAR ON A PSYCHOLOGICAL THEORY OR EVALUATE A MODEL--(2) A MODEL IS PRESENTED TO INDICATE HOW INFERENCES CAN BE MADE BY CONSIDERING A FIXED MODEL ON VARIOUS LEVELS OF DATA ANALYSIS AND IN VARIOUS EXPERIMENTAL SETTINGS ("ALL-OR-NONE MULTI-LEVEL MODEL")--(3) A MATHEMATICALLY RIGOROUS FRAMEWORK IS DEVELOPED FOR ANALYZING A LARGE CLASS OF MODELS WHICH EMBODY ITEM DEPENDENCIES--(4) THIS FRAMEWORK IS APPLIED TO SPECIFIC MODELS (INCLUDING THE "ALL-OR-NONE MODEL") AND RESULTS FOR VARIOUS PRESENTATION SCHEDULES ARE PRESENTED TO ILLUSTRATE THE FLEXIBILITY OF THE FRAMEWORK. ALSO ANALYZED IN TERMS OF THE FRAMEWORK ARE MARKOV MODELS DERIVED FOR A PARTICULAR CHOICE OF PRESENTATION SCHEDULE, AND RESTLE'S STRATEGY-SELECTION THEORY. THE FINAL CHAPTER REPORTS ON TWO LIST-LEARNING EXPERIMENTS CONDUCTED BY THE AUTHOR TO GENERATE DATA RELEVANT TO THESE IDEAS AND METHODS OF ANALYSES. (JD)