A study was conducted among 64 students in an introductory psychology class at Yale University to clarify the components of syllogistic reasoning used in a syllogistic evaluation task. A modified form of componential analysis was used to decompose the syllogistic evaluation task with abstract content into the subtasks of encoding and encoding plus combination. The response-choice data obtained from these subtasks showed that more direct tests of various models of syllogistic reasoning were derived than had been possible, and in particular, tests of assumptions about sources of error in syllogistic reasoning. More direct inferences regarding the representation of relations between the subject and predicate as encoded and combined were also provided. The study supports the superiority of the transitive-chain model of syllogistic reasoning. References, Euler diagrams of set relations, and seven tables of experimental data are appended. (Author/DF)