Previous research conducted from a schema theory perspective has neglected those instances where the reader is confronted with unfamiliar text. This issue is addressed, and initial results on training students to acquire and use form schema in processing relatively unfamiliar descriptive prose are presented. Thirty-two college students were randomly assigned to two groups: the DICEOX group which received training on the use of knowledge schemata and the control group which did not receive training. A 2,500-word passage dealing with the theory of plate tectonics was used as the material to be learned. The Group Embedded Figures Test (GEFT) and the Delta Vocabulary Test were employed as measures of individual differences. Dependent measures consisted of an essay test, short answer test, multiple choice test, and cloze test. A 2 x 2 analysis of covariance was conducted for each of the four dependent measures. Results revealed that: (1) the DICEOX group performed significantly better than the control group on the essay exam measuring comprehension and recall of descriptive prose material, but did not perform better on the other three dependent measures; and (2) high scorers (field independent) on the GEFT significantly outperformed low scores on all the dependent measures. (Author/RL)