To ascertain if children's use of manipulatives and/or the teacher's presentation of the problem in parts (chunking) affect children's ability to model and/or solve addition and/or subtraction problems, 146 first-grade subjects were administered the Houston Addition and Subtraction Problem-Solving Test which provided accuracy and modeling scores for two addition and three subtraction problem types. The test was presented in holistic (not chunked) and parsimonious (chunked) forms. Results indicated that chunking significantly affects children's ability to solve nontransformational addition problems. The interaction of chunking and modeling significantly improved the subjects' ability to solve take-away subtraction problems. (Author)